 Great. Okay. Cool. So thanks again for the chance to speak at the Berkeley Institute for Data Science. I just wanted a quickly level set to make sure that we're all thinking about data science in the same way, both outside and inside the Institute. So as you've heard, I used to be at Stamen, now at Code for America. Stamen Design is a design firm. We, well it still exists but I'm no longer there, but for the eight years that I was there we were doing some of the early work in information, visualization, mapping, cartography, all kinds of ways of doing visual communication of data on the internet using sort of live API based, web based data. That was fairly a new thing at the time. So a number of projects we've worked on really early on about 10 years ago or so. We did something called the moveon.org town hall, which was one of our first forays into doing stuff that was live with data live on the internet. This was a project with moveon.org, which if you remember in the 2004 election cycle was kind of one of the main sort of lefty political action committees that was operating at the time. And they were really interested in using this enormous kind of web based platform that they had to have these kind of house party meetings around the country and get feedback and kind of, you know, participation from people around the country in preparation for the 2004 election. So we developed some early work in online cartography showing this kind of live feedback of people. This would be accompanied by an audio transmission, essentially kind of like a radio show. And then all these different dots and patterns would come up in response to people's interaction with the show. So the show would kind of put out a question like how many of you are, you know, going to participate in a phone party or how many of you want to go out in canvas for health care or something. And this would give you a live dynamic visualization at your house party showing what that response was like. So fairly early kind of abstract visualization work. More recently, around the time that I left in 2012, we were doing some sort of kind of more visual textual work, textural work like the watercolor maps. This is based on an open data set called OpenStreetMap, which is a worldwide collection of streets that have been crowdsourced by millions of volunteers around all over the earth that are doing kind of building and street and bridge and other kinds of tagging. So we did this project called Watercolor Maps where we applied a kind of digital procedural watercolor textural kind of look to them that people really kind of caught on to as an interesting way to visualize geographic data. So that's Stamen. I spent about eight or nine or so years there from 2003 to 2012. And then more recently, I've been at Code for America. Has anyone here not heard of Code for America? One, two, okay, a few. So Code for America, we're a nonprofit. We're based in San Francisco just across the bay, but we work all over the country. We're on our fifth year right now. And our focus at Code for America is really in helping kind of to make government better, to make government work using the digital principles of the 21st century. I'll start by just kind of showing a few projects and then kind of get into some of the work that we're doing and the reasons why we're doing it. So Oakland is a city just to the south of here. It happens to be the one that I live in. And in 2013, we worked with Oakland to produce a record track. It's currently available at records dot oakland.com. And it really kind of addresses this problem that Oakland was finding where because they didn't have a mature set of digital tools and a mature way of communicating digitally inside the city, things like public records requests were really starting to fall through the cracks. In particular, in 2013, the problem that Oakland was facing was a sort of kind of just nasty, perfect storm of kind of immature digital tools, different departments that were using the public record system as a way to kind of offload work onto one another, and then a sort of unprecedented increase in just the raw number of public records requests that were coming out as a result of the Occupy movement. A lot of people really unhappy with the city's handling of the Occupy protests on City Hall Plaza. A lot of new public records requests dealing with things like police treatment of protesters, things like that. So they were just really caught in a bind. It was a really bad situation for them. So we produced this piece of technology called record track as a web project. It's a very simple one, you know, sort of a basic web browser based interactive environment where staff and citizens can come together to see the progress of open public records requests, see the result of other people's open public records requests and also make their own. This has become hugely popular with people like, you know, real estate agents, other folks that often have to deal with things like city records and so forth. So it's become a really nice project that we kind of point at to show ways in which city operations can be helped with digital means and digital technologies. More recently, last year in the city of Atlanta, we did a project for the Atlanta Municipal Court website. So previously it had a fairly primitive website. Post this project, you could actually get information about your own court case. I'm going to skip over this one right now because I actually have a lot more detail about the Atlanta work a little bit later in the talk. So these are just kind of two examples of the work that we do at Code for America and the ways in which we help our partner governments, our partner cities over the course of a one year fellowship program. So this year we've got eight of these things going simultaneously. There's 24 fellows out around the country right now in cities like Richmond, Virginia, Somerville, Massachusetts, Vallejo, California, et cetera, kind of looking at digital problems and database problems that these different municipalities have and attempting to come up with some kind of design and technology solution to those problems. Our mission statement is this. Government can work for the people by the people in the 21st century. There's a lot of stuff baked into that statement, but I think that the 21st century bit at the end is really the one that's meant to highlight the fact that we're working on digital technologies and digital communications. We're kind of in a place right now historically where digital stuff has started to become just the basic way in which people interact. You know, the same way that like wireless sort of started to get dropped from communications and horseless got dropped from carriage eventually. I think we're going to see the word digital start to disappear over time because it's just kind of the way things happen now. And cities in particular, the municipal governments that we work with, both city and county, are in a uniquely bad position here because they've been dealing with these kind of budget pressures and other kind of unfortunate input over the last years that puts them in a uniquely bad situation to be reacting to this huge change in people's expectations. A shorter way of saying what we do is that we make things and we make friends. We make things in the sense of technology, we produce work, we do design work, we do research, we do, you know, deliverable pieces of code and work for these cities. But at the same time we have a very large communities program because I think, you know, really early on our founder kind of cottoned on to the fact that you don't just change this stuff at a kind of purely operational behind the walls level, you also have to have some kind of community that connects this stuff. So we have a program called the Brigade Program. That's in addition to the Fellowship Program. And that's sort of for those people that don't want to be, you know, involved in a full years project, but they maybe just want to, you know, live their lives, be in their town, but show up at City Hall maybe once a week for a hackathon or show up at City Hall once a week for some sort of meeting with city planners or other city people in order to figure out how they can help locally, digitally. There's over a hundred chapters around the country right now. If you guys are interested in the Brigade Program, we have two local ones. There's Open Oakland, which meets on Tuesday night 6.30 City Hall and Code for San Francisco, which meets at our office on 9th and Natoma in San Francisco on Wednesday nights. So if you're interested in any of the stuff that I'm talking about today and interested in volunteering, those are two great groups to come and join and hang out with. A lot of good folks. This is a chart that Stuart Brandt put in the Clock and Along Now book and it, he's talking about this idea of pace layering. This idea that there's sort of different layers to like nature, culture, governance and infrastructure that move at different speeds over each other. And I think the work that we did at Stamen was in large part focused on the kind of commerce infrastructure fashion, kind of fast moving visual design stuff. And at Moving to Code for America, I realized that it kind of gave us a unique opportunity to really think about these two layers. Governance in the sense of policy and the rules the government has for how people need to interoperate and infrastructure in the sense that we're helping these governments kind of influence their governance and their policy through the creation and delivery of new tools and new forms of doing digital communication. This diagram is a very kind of clean and neat one with these clean and neat divisions. But I think one of the things that we've really discovered over time working with these governments is that these layers tend to sort of bleed into one another and feed into one another in some pretty often unanticipated ways. There's a sort of more engineering type example of this that I have often liked to refer to. This is City Court Plaza or 601 Lexington in New York City. And for those of you who've seen the top of this building in like sky shots, you know that it's got this really kind of like distinctive top of the building. But you might also have known that it's got a really distinctive bottom of the building as well. The engineer who produced this called it a skyscraper on stilts. I believe his name was Le Moussier. And the reason why it's constructed like this is because they had to deal with this kind of governmental slash like tax parcel land ownership situation with the church that's down there in the lower left hand corner. I forget the name of the church but the church basically said look we're not going to move our church just because you want to build a skyscraper here. So they constructed this thing on these funky stilts with this huge kind of sky frame around it in order to leave room for this thing. And it's only come out in like about the past 10 years or so that when it was constructed in 1977 the engineer who constructed this thing got a piece of feedback from some like undergraduate engineering student someplace who had done a little bit of extra research on this frame and a couple of new simulations of some sort basically showing that this whole thing was going to collapse in the event of a particular kind of hurricane. They did the math, they figured out that this particular kind of hurricane, these particular kinds of what's called quartering winds were expected in New York about once every 16 years or so. So there was like a really real chance that this whole thing was going to collapse on top of somebody. They had this you know this like nature layer kind of bubbling up through infrastructure and really kind of kicking them in the butt. The story came out like 20 years after the fact that they actually ended up doing secret undercover engineering work on this building while it was already occupied by tenants over the course of weeks and months in the 70s basically sneak while not sneaking in but getting into this building late at night exposing small parts of the frame through the walls internally and strengthening them specifically so that it could deal with these kind of quartering winds. So it's this kind of amazing example of nature just punching through these layers and like laying the best laid plans of mice and men a waste. We see this in our work not in structural engineering but in things like statistics and the way that they actually deal with regular kind of people on the ground. This is a chart from New York Times late last year connected to the Ferguson protests that were happening basically last fall and this is actually this is a statistic about arrest warrants and what's called bench warrants. This is showing the number of these warrants that have been issued per citizen for a number of different cities throughout Missouri and you see that it's got this kind of characteristic hockey shape hockey stick shape to the right and what it's showing you is that for every thousand residents in Ferguson during this time period in 2013 1500 warrants were issued for these folks. So one of our funders from the Arnold Foundation actually brought this number to our attention about two or three years ago and said hey you have to be looking at this stuff like there's this data out there that's showing kind of a brewing ugly situation in places where a police department feels that its citizens are like a source of money to be shaken loose rather than people that are actually being served. So our funder kind of brought this to our attention and basically said look like if you guys are thinking about data science and these projects and open data and all this stuff this is the kind of impact that it's going to have. It might help you look at you know Jennings or St. Anne or Hazelwood or these other at risk cities and help kind of deal with these problems before they actually erupt in the way that they did in Ferguson. So there's sort of three ways that I think about our work at Code for America and that's that we're learning under constraint under the really sharp constraints that city governments are under that there's sort of things that government knows the government often works with these kind of known unknowns and very typical ways of dealing with patterns and then how government copes at the unexpected and how we can start to see that kind of from a little bit longer distance and help governments predict these things. So in terms of learning from constraints this is basically the situation that we walk into in like every city government or municipal government that we have contact with. They're all in a really bad way budgetary budget wise. City of Oakland tells us that something like 15 years ago was their high watermark and ever since then they've been having to like clip the budget by some amount every single year. So this manifests in the form of hiring phrases manifests in the form of inability to you know pay for new infrastructure and really kind of ultimately manifests itself in the form of government not really knowing how to learn and not really knowing how to do stuff. This is a federal example but I think it really illustrates the problem. There used to be something called the Office of Technology Assessment which worked with Congress and right around the time of the mid-90s sort of contract with America time period Newt Gingrich time period we saw things like the Office of Technology Assessment have its budget completely slashed out from under it. This was the office that would have looked at something like the healthcare.gov debacle and said okay wait a minute there's something wrong here you know we know something about technology we know a little thing or two about assessing it and we can see that there's some sort of situation looming here on the horizon they need to fix. So when something like the OTA gets kind of ripped out from government it saves the money that's in the budget for the OTA but that savings kind of gets transferred into situations where because government no longer knows how to assess reality and technology it starts to lose the ability to predict things. The other kind of big impact that happened with cities is the kind of series of recessions that started in 2001. My friend Kellen is Chief Technology Officer at etsy.com they're an online goods marketplace that sells like handmade crafts, antiques, things like that. He spent a lot of time working at companies like Flickr and others that are sort of we're kind of at the vanguard of new ways of doing internet technology stuff what we would have all called the web 2.0 about five or six years ago. And he talks about how there's this interesting dynamic with a recession where even though the money goes away and ostensibly you can't pay for good stuff anymore you have this kind of he calls it a classic post-industrial creativity zone. You have these people that are smart and interested and want to do good stuff and really just need to kind of like pay rent or a mortgage and food and so forth and are motivated at that point to go and do interesting stuff. So coming out of the 2001 recession you have the kind of post.com boom money just kind of went away and what was left in its place was actually some of the most interesting currently kind of huge and powerful technology movements that we've seen happen in the past 15 years. Facebook kind of comes out of this environment. Twitter comes out of this environment. Flickr comes out of this environment. All of these currently huge companies that had to suddenly deal with a huge constriction in terms of what they could do financially and responded to it creatively. Two folks at Flickr developed something called dev ops or dev operations that's Paul Hammond on the left and John Alsbal on the right looking lovely. And they were starting to look at things like how you do operations for technology in a big company. You know they happen to be working at Yahoo which on Flickr at the time. And they were starting to see ways in which you could kind of take job descriptions and recombine them instead of having let's say developers in one camp and operations people in another. If you started to combine those job descriptions and combine those teams and get them talking to one another. You weren't just saving the money of not having to pay as many engineers but you were actually getting a particular kind of creativity coming out of the fact that these folks were no longer in an adversarial relationship but we're now in a more kind of collaborative relationship. They were developing these systems together and thinking about both the engineering and the code and the operations and kind of how it would actually play out in terms of like physics and machines and hardware and speeds and so forth. John likes to talk about you know these charts that he started putting together at some point where they began to think about Flickr not in terms of like cost for hardware but like number of days of Flickr left. They essentially had these charts that were showing like at current rates of growth when would they run out of things like hard drive space or memory or other stuff and they started to develop this really kind of tight sequencing feedback loop based on the data that they were seeing and based on job descriptions like these that allowed them to sort of save money and get really creative in terms of how they did operations. That led to things like the cloud. You know Amazon all these other big companies that are starting to think about software as something that lives up in space rather than something that lives on your desktop. So for cities this is a really big deal. You know a lot of the governments that we work with are still I think stuck in a fairly 90s mentality about their technology. They think about stuff as being servers on premises or things that you have to like buy and install someplace pieces of hardware. And because they can't think in these terms or they don't yet know how to think in these terms they end up really kind of wasting a lot of cycles thinking about things in fairly old ways. You know you've got email at the city of Oakland that goes down for multiple days at a time where if they were using you know Gmail or Hotmail or something else it just wouldn't be a problem. And there's like people whose heads roll in these departments instead of it just being a purchasing decision. So in that recession you've got this kind of Cambrian explosion of all these different technologies. And a lot of the work that we're doing with these cities is really trying to focus on how do you get this explosion of these technologies into a place where cities can actually start to take advantage of them. So I'll talk about two examples. San Francisco and Atlanta in San Francisco in 2013 we did a project with the Human Services Administration there. So essentially HSA deals with things like administering food stamps or housing the homeless. You know all these kind of like very core like day-to-day survival kind of operations that have to happen with the poor in a city like San Francisco. Our team that was working on that continued to work at Code for America in the intervening year and they've produced a number of applications using the Twilio API. For those of you who are not familiar with Twilio, it's a corporation who produces basically APIs on top of SMS text messaging and voice phone recording kind of interaction trees. So you can do things like you know produce without having to buy any hardware of your own an API and an application that actually deals with the phone system in a way that you couldn't before and you pay for it like pennies at a time instead of in the form of giant capital expenditures. So they've produced two applications based on the Twilio API. One is called Balance which seeks to address this issue where they were finding that welfare recipients were paying these huge amounts in aggregate of ATM fees in order to interact with their cards. You know to do things like a balance check or to do things like take money out of your EBT card. You know it costs you like a buck here and a buck there but it really adds up and it really felt like the banks were kind of taking advantage of this. So they produced an application called Balance which very simple. Essentially all it does is it kind of opens up an SMS or text-based API to figuring out what your balance is. But it addresses this really core critical problem that the users of these cards had which was that because a balance check was expensive or difficult or time consuming or required a long phone call you know you'd show up at the store thinking you had 10 bucks on your card and then get up to the front of the line at the register and find out you only had five bucks and you were embarrassed. You felt awful. Your kids didn't get to eat that night. Like it was just a really bad situation for a lot of these folks. So we're finding that doing these very tiny interventions around applications like this allows people who are recipients of these you know these public welfare programs to have a more humane and kind of more sane interaction with them. Another project that the same team built this year was called Connect. Same kind of situation. You've got these low income families that are waiting on hold for hours and hours every single year just trying to kind of get information about like what the situation of their cards is. Connect is very simple. You try to you know get on a phone call with one of these representatives and essentially a computer does the waiting for you. It's very simple. You get a phone call back saying 45 minutes have passed. We now have somebody on the other end of the line who can actually talk to you and you've saved that 45 minutes without having to like sit there like waiting for this thing. This is actually a really fun one because it got the attention of the director of California Social Services Will Lightbourne and he actually lent his voice in a recording for the application so that when somebody who's like a you know phone operator at one of these services picks up the phone it'll be a recorded voice of Will's saying hi you have a client on the line and they kind of stiffen up and realize oh shit okay best behavior this is for real. And so it was kind of like a nice kind of like voice based phone technology hack that removes this kind of small but very consistent pain from the lives of these citizens. In Atlanta the courts project was sort of a similar kind of basis. In 2014 our fellows there found that the Atlanta court system had this kind of really unfortunate interaction style with the citizens. You know if you got like a parking ticket or something and it told you to show up on a court at court on a particular day this is the kind of thing you were faced with when you went to court. You know line outside once you got in the building you were waiting for your name to pop up on a monitor. Then you'd you know kind of stand there like wondering what to do next then there was a line for the elevator then there was a line once you got upstairs to court then there was a line to pay your ticket. It was a really unfortunate kind of series of these like micro abusive or micro aggressive interactions that the fellows thought that they could do something with. And again it's you know they're looking at things like phone technology and SMS technology and sort of simple small operational hacks that would manifest themselves in a forum like this. They got police officers to hand these little pieces of paper out to all the different people that were interacting with the court system through things like like parking and traffic citations and basically presented them with a telephone number that they could use to text them to figure out if they even had to go to court. So you know of the 200,000 people that were cited annually in Atlanta we were seeing numbers like 40,000 people miss court. So as a result of this work in Atlanta we're seeing 10,000 people checking on their cases. Maybe they don't have to go to court after all or maybe they are getting reminded when they would have forgotten otherwise. It's sort of a small way to help people understand whether they actually have to go and wait in that line and hopefully prevent certain people from going and kind of plugging the line if they don't have to be there anyway because they can pay on line. Interaction wise super simple you just text your number in them and you get a response and you can interact with this thing using this kind of text based like kind of side channel which allows you to suddenly have this much more you know kind of sane level of interaction with a government program that would normally be like phone calls and lines. So you know small stuff small technologies but things that government doesn't believe themselves to have access to when they talk to their normal vendors about things like SMS or voice they get quotes of you know 30, 40, 50 million dollars from the Xeroxes of the world who view this stuff as a giant IT project rather than some sort of small hack that can be kind of expanded into something that works. Second way that we think about stuff is kind of managing known unknowns. Government sort of traditionally deals with kind of big problems. This is a photo of Oakland about 100 years ago. San Pablo in 14th it's actually where City Hall is now and you can see so much government infrastructure cable cars and roads and all the kinds of things that government has to do. And I think when government thinks about the kind of problems that it needs to solve typically its mindset is very much sort of you know Mr. Wolf if you remember the movie Pulp Fiction. I solve problems if I'm curt with you it's because time is a factor. For those of you who remember this movie you'll recall that you know Mr. Wolf didn't come in with any kind of like split the atom type solution to folks problems. He was just like all right you got to clean up a dead body in a car. Well you know you get some towels you get this you get some hefty bags give me a cup of coffee like lots of sugar lots of cream really basic really straightforward really kind of known known way of approaching problems. And one of the things that we've seen as we work in government is that they're kind of more attuned to that way of thinking but they're not always attuned to the kind of unknowns that come out of technical problem solving. In kind of engineering circles people refer to this as waterfall versus agile and lean problem solving where waterfall is a very kind of linear way I shall write you know a request for proposal for a system that will solve this problem for me. People will bid on it I will choose the lowest bidder and then we're going to go ahead and run with that. The things that we're finding with digital communication is that methodologies like agile and lean from software development are starting to become a much much more important way of thinking about these problems. For those of you not familiar with agile lean it can be really summed up this way. You have an idea you build it with some code you measure whether it works you get some data back and then you learn whether it actually made any sense and then you kind of go back to step one. It's pretty basic stuff it's just you know it's an iterative way of problem solving where rather than thinking of things as a kind of A, B, C, D, E sort of step following exercise you think about them more as a learning process and this becomes really important with digital tools because they change so quickly. You know five years ago there wasn't I don't know iPhones in the world and now there are or ten years ago people had different expectations of computer speed and now they have you know much faster expectations of computer speed. So it's kind of all of this digital stuff that's happening at a speed the government is not typically attuned to but it kind of needs a way of figuring out how to operate in this way and that's what we think we're helping them with. So just a step back in time for a second to a project called Crime Spotting which I actually did when I was still at Stamen Design. You know this is kind of my first exposure to government technology. We were looking at crime in Oakland in particular and it was sort of spurred by an interest in it that resulted from the assassination of a journalist just run the corner from my house. I don't know if you remember the editor of the Oakland Post but he was basically shot in broad daylight like one morning as I was going to work. So we kind of got interested in crime at that point and realized that Oakland's method of publishing crime information was something called Crime Watch which was very kind of directed. You have a question. Was supposedly I think convicted of the crime. Yeah. Yeah, Chauncey Bailey I think was his name. Yeah, so the Chauncey Bailey case that was around the corner. That sucked. So, you know, Oakland would publish a thing like this and in order to even get to a screen like this where you have some sense of what kind of crime is happening around you in your neighborhood. You had to step through this gauntlet of screens like what neighborhood do you live in? What types of crimes would you like to see? What time period would you like to see them from? So it wasn't like an exploratory interface. It was much more of a like, you know, known, known kind of like please enter the info you would like to see kind of situation. We built a project called crimespotting.org. The interface, you know, it's still dots on a map fundamentally, but the idea was that you could go into this thing and see them all at once and then start to narrow your selection based on what you see rather than narrow your selection based on something you don't see. So a lot of really basic interface elements here. You can look over time in this interactive way. There's that thing called the pie of time in the lower left-hand corner that allows you to look at different times of day. For example, if you're passing through downtown typically in the evening, maybe you only want to see what kind of crimes are happening during the times that you are there. We thought it was a big improvement. These were the icons that were in use on the map in Oakland. You know, some of them more suggestive than others. I think the middle one was like a syringe for all types of drug crime of any kind. This sort of boogery looking thing is a knife. The big pink pea, I think, is where they chickened out a little bit. That's actually for prostitution. They could have been a little bit more representative, but I guess they didn't want to. And we had a lot of interesting downstream projects from this. You know, once you have this kind of data, you can start to produce things like heat maps. This is the thing we did trying to illustrate where the hotspots were in Oakland crime. So predictably, you can see this kind of 14th San Pablo Telegraph kind of intersection. It's really the heart of Oakland downtown is where a lot of the like small petty thefts and things occur. But then there's also blobs to the north that you would look into and find that there's, let's say, a lot of drug crime or prostitution or just like really hassled cops up there or something. So this kind of gives you like one element of this where you're looking at this open data that's coming from the city, showing it in a different way, providing citizens with a way to see that stuff. But the thing about knowns and unknowns and known unknowns is that you have to start to kind of close that loop a little bit like in the agile way, start to have there be a way for citizens to actually feed back into this thing. So we did a project in South Bend in 2013 that we've been continuing since then called City Voice that at Code for America attempted to kind of close that loop somewhat. So in South Bend, the problem that that city came to us with was basically vacant to the abandoned properties. It's a small city, kind of rust belt, Northern Indiana, I believe. They had huge numbers of abandoned properties. This is especially bad in the wake of the foreclosure crisis, the financial crisis. And they just didn't know what to do with these things. They knew that they had like thousands of houses that didn't have owners in the city. But as a city, as a government, they didn't really have an understanding of what they were looking at. They didn't know if they should all be bulldozed or if they should be sold or what. So our fellows created this project called City Voice, which for citizens kind of provided this sort of interface. They actually put together signs and assigned call-in codes or property numbers to every single individual property and stuck them all over town on the front yards of these properties. So the theory was that rather than trying to get people to come to some website or rather trying to get people to visit something on their phones or download an app, they would just be walking by and thinking, man, that house is really messed up. I wish somebody would do something about it and have a number where they could actually call in right at that moment and offer some opinion on it. What the city was looking for was some sort of distinction. You know, they knew that some percentage of the houses were just like needed to be torn down. Some percentage of the houses were probably fine and could be sold for money, but there was this big, fat middle section where they didn't really know and needed to hear from local residents to understand what was out there. If you visit the site today, it's southbenvoices.com. You can see an interface like this where you can actually play live recordings from members of the public that have called in and said, you should tear it down, or I think it's a great house and I would buy it. So it's a really important way for the city to be able to kind of start to see this feedback from the data that it puts out there. It puts out data about where these vacant and abandoned are and the city comes back with information about whether they're actually worth saving or not. It's a really big deal because it completes that loop and gives the city a way to hear not just kind of an abstract like push one or push two type opinion, but the actual voice of a human citizen that has a feeling about this particular building and lives in the neighborhood and wants to see something good happen to it. Yeah. Blight status. Yeah, there's a relationship to it in the sense that they're sort of similar projects and informed by some of the same considerations. Blight status was a project that we did in 2012 in New Orleans. Very similar concept also vacant to the abandoned properties, but it was built not around this kind of telephone means of interaction, but it was it took advantage of the kind of stronger citizenship like clusters inside New Orleans where they found particular people, you know, oftentimes like older folks that had lived in these neighborhoods for 60 years who were more interested in kind of seeing this data to sort of provide directed feedback. Like I've lived here for 60 years and I know the history of every one of these places and can provide you kind of bulk feedback in a kind of citizen, you know, like town hall meeting kind of kind of structure. But yeah, very similar thread runs through these things. I think city voices was directly influenced by the work on Blythe status. In Oakland, yeah, I think you might you may be talking about the soft story project, which I'll talk about it like towards the end. Yeah. So I mean, it's really similar stuff. I mean, the thing about these projects is like if you kind of zoom out from the work of Code for America, it sort of starts to all kind of blend into itself because it's really the same set of problems or overlapping sets of problems in all these different places. And what's unique about each one is more often a kind of administrative like gridlock in these places rather than any particular identity of the problem itself. So the kind of last way that we think about this stuff is in the form of like surprises from below, like what happens when you start to get these feedback loops going and those lower levels of, you know, nature, culture and government start to bubble up. For those of you familiar with Maslow's hierarchy. Sorry. There's something we came into contact with via the healthcare.gov rescue effort that we've been calling Mikey's hierarchy. Mikey is Mikey Dickerson. He's part of the team of a half a dozen folks that worked on the healthcare.gov rescue. And, you know, when he talks about his work there, this is sort of a summation on his mind of what he was doing. For those of you who can't read it, it's a pyramid. It says monitoring incident response, postmortem, testing, development and product. And his argument when he kind of showed up at healthcare.gov and realized what a mess that it was, was that because nobody was doing, like actual monitoring of whether the service was working, it meant that they couldn't have discussions about any of the higher level things. You know, literally initially their monitoring was CNN. Like they didn't know if the system was up until either somebody tried to sign up for healthcare and failed or somebody went on CNN and said, it's broken. This thing is stupid. And so a lot of the early work of the healthcare.gov rescue was literally just people who came from, you know, Google and other types of companies that are extremely mature in terms of thinking about this kind of operational infrastructure and applying it to a new environment where literally no one even thought about this. The way that Mikey describes this is that they didn't have outages before because an outage has a beginning and an end and they just have this kind of constant rolling low level crisis of not having any idea what was going on. Once you have outages that have a defined beginning and an end point, you can start to make them shorter or longer through your actions. But you can't actually influence any of this stuff until you know what's out there. And so literally you can't talk about product and development and your ideas about how to help people until you get all this shit sorted down here. It's something that an architect friend of mine, Brian Boyer, calls the matter battle. The conflict between human intentions and the laws and behaviors of the physical universe. Bunch of different characteristics of the matter battle. It's hard to predict matter. You don't really know what it's actually going to do. This kind of like ties back to the city corp tower thing. You don't really know what it's going to do until somebody runs a really good simulation or you actually kind of get in there and try. There are always low tech. It's really low level stuff. And it's hard to understand what is actually going on and kind of what situation you're in until you're actually in a matter battle. He says that we've been to the bottom of the ocean on the top of the heavens. Putting things together, taking them apart rarely goes exactly according to plan. Obviously with computer stuff and digital stuff, we're not really dealing with matter. But there's other forms of matter that are in that environment. This is a quote from Tim Gregory, who's the head of CGI British operations. For those of you who remember CGI Federal is the international company that worked on the original healthcare.gov and by all accounts kind of bungled the process. And this is a really amazing quote. He was actually allowed himself to be quoted saying this in a major national newspaper, which kind of blows my mind. He's essentially copying to the fact that a company like CGI Federal operates at this purely kind of lawsuit driven development level where whether you're successful or whether you're a failure is really more based on whether you can sue the other guy or they sue you more so than if you're like actually building technology that functions and works for people and helps them. Like this is how a company like CGI Federal views its work compared to somebody like Mikey Dickerson who's like well are we saving people's lives or helping them get healthcare? Like what are we actually doing here? And it starts to when you start to think about like you know data science and these kind of other like you know sort of geographic and data based kind of digital communications approaches that an organization like CFA gets into. It starts to touch on this idea of like you know who can actually see what's going on on these projects. There's a fascinating book. I don't have any of you read this book by James C. Scott seeing like a state and if you've not read this book please read the book. It's so good. It's so it's so deeply good. Basically it's this book length argument. I believe the subtitle is something like how certain attempts to improve the world have failed or improve the modern condition of failed. It's like a book length argument against like state level statistical reasoning about the world. And a lot of what Scott is pointing to is this idea that a government needs to see things in an ordered fashion like some parcel of land needs to have one and exactly one owner. You know some chunk of a farm needs to have one thing that it produces where in reality you know especially in a lot of these like kind of pre-modern civilizations like I don't know Europe more than 200 years ago things kind of operated at a lot of different levels. You would have farmland that would simultaneously provide like you know I don't know honey bees or animals for hunting or other things. And a lot of the ownership patterns and the kind of stewardship patterns around this stuff were very much kind of locally negotiated. And it wasn't until like the tax official or the land speculator got involved that you had to start seeing things in terms of this governmental kind of you know spatial geographical way of way of modeling. You know government is sort of the keeper of space in a lot of different ways that we like to think about. And it's sort of interesting to be a code for America because on the one hand we're sort of we're beneficiaries of this incredibly high level statistical processing apparatus like the U.S. Census or the U.S. Geological Survey. But at the same time you have to kind of maintain a certain level of awareness of about like when it's going to pop up from you know below the floor and kind of kick you in the butt because you've been making the wrong assumptions. This is a great slide. This is this is the state plane coordinate system. So these little dividing areas are basically like a series of geographic projections which let local engineers pretend that the earth is flat in those particular spots. So we're over in California zone three over there and it's a you know a chunk of counties that goes east to west which allows you know basically like civil engineers working in California to figure out that bridges can meet in the middle even though they start in two different counties at either side. If you go down into demographics same kind of stuff. This is a screenshot from a Stamen project we did a number of years ago called this tract dot com where we were just looking at like individual tracts out of the U.S. Census and trying to figure out what it was that the census actually knew about all these places. So you know really basic stuff like this is block number 25017353700 blah blah blah and the census will tell you that because it's got Harvard University sitting in the middle of it it's got like an absurdly high level of college graduates living in there. And at even further down level you've got this kind of atomic unit of you know human land ownership. This is a parcel data set. You know every tax assessor in the country has something like this has these plat maps that show basically who owns what where and who you bug in order to like pick up tax revenues from that place. It's how government funds itself is by looking at land in this way and dividing it up into these kind of neat squares for a company like Stamen where we were really interested in the sort of you know the kind of algorithmic and data exhaust of this stuff it led to some pretty cool projects over time. We worked with trulia.com which is a kind of real estate site really early on. This was probably like 2008 or 2009 or so. They gave us these huge data sets that we had never seen before of like individual buildings and their histories and what they knew about them. So we did really basic stuff you know like this is San Francisco with these dots animated to show construction dates over time. So you can see this kind of bubbling here in the 2000s but it's preceded by a very slow rise and then like a sudden sort of east to west flood of people filling in the sunset district. So you get these kind of like gorgeous very high level visualizations but I think what what Scott's argument and seeing like a state is is that this sort of stuff tends to hide a lot of lower level stuff. You know if you're looking at things at this kind of parcel or tract or country level it's really easy to miss the fact that there's stuff that's happening way way way down at the bottom that might be either a help for you or might be a threat for you. Soft story I think is the Oakland project that maybe you're thinking of. This is a project that the open Oakland Brigade which meets Tuesday night City Hall 6 30 p.m. has been working on for the past couple of years where they're looking at this particular form of data around a construction method called soft story construction. Anybody never heard that term before? It's a reference. Yeah, a couple of people. This is a soft story house. I think this might be somewhere east of the lake in Oakland and you can see here that it's got this kind of open bottom story. So you know these things open up cars drive in. You can imagine there's not a lot of walls or major structural elements down there because you need a little bit of room to like have a store or park a car or something. These things are like crazily dangerous in an earthquake of the sort that we tend to have around this area. So there's been this data set that one of our former fellows Dave Garino has been working on to try to get out of the city of Oakland and the County of Alameda around where these soft story buildings are and what their status is with respect to you know civil engineers inspecting them and people sort of you know buttressing them and doing like kind of basic you know structural work to make sure they don't fall down and pancake people. So we produce this this work with open Oakland. It's soft story dot open Oakland dot org. And you know it's it's basic right. It's just it's a map. It's kind of like crime spotting. It's kind of like any other dots on a map kind of application. And it shows you these sort of you know red marks that tell you that there's a potential soft story building there. Doesn't actually mean it's going to fall down but it means that there's a risk that it's going to fall down in comparison to the green check marks which show you where there's buildings that have been you know reinforced and have been inspected and proven to be you know up to a higher standard of earthquake construction. So you know on one level it's kind of this cool sort of you know data science map of what's going on. But at a deeper level there's a whole series of questions that starts to open up around things like you know poverty and access to expensive construction and the relative rents in these places. Like if you go into one of these spots and reconstruct it so that it doesn't fall down on an earthquake because that in some family has to get booted out. And that was the only place that they could actually afford to live. So you know a lot of the work that we do is working with cities basically not just to kind of like you know beat data out of them and put it online but to also be in a dialogue with them to sort of understand what their concerns are about how to get this data out in the public. Like Dave Garino spent three or four months working with different government employees to get this data set because they were deeply deeply uncomfortable about just dumping it online. They wanted to see it sort of treated in a way that the language actually referred to the kind of truth as they saw it. So it says potential soft story buildings in Oakland Oakland on top. It doesn't say soft story buildings. It has to kind of soft chew a little bit and say OK there's a certain like statistical likelihood here that we're dealing with not just like a thumbs up thumbs down building. And that starts to kind of tie back into this Ferguson chart. You know if you can start to count things that are that are affecting people at this low level then potentially you can start to see these problems coming from a little bit further afield and deal with them in some sort of way before it actually becomes a crisis or emergency or a boiling point. So we have this like strong feeling that government actually must work. Like you know if you go back to this pace layering chart like everything else sits on top of governance everything else sits there and it's resting on that level and if it doesn't work then like whole other chunks of infrastructure and commerce kind of fall through down to culture and nature. And you know that that could get pretty ugly. It's also just a fact that you know for a long time at CFA people have sort of talked about us as these kind of you know bringing innovation to government sort of folks. We've started to kind of back off from terminology like that as we've been thinking about things in this layered way. Because in a lot of cases innovation isn't really what you're looking for. Amy Hoy had a great quote about a book that she's putting together saying fuck innovation things we use every day are so disgustingly terrible that there's unending opportunity. It's a sense that like you know there's like so much opportunity to fix basic stuff fix basic interactions get basic data published really low level operational stuff that you can just skip over innovation entirely and head straight for operational things and actually help regular citizens with you know bench warrant counts and soft story buildings and you know feedback about all these are baking the man and properties and so forth. So I've got half an hour for questions. Thank you. Just to let everybody know if anybody has to read together class or work or something we do have an extra half hour for questions and this is being recorded and because we don't have a mic to pass around right now Michael repeat your questions. Michael will repeat your questions. Yes. What prevents the IT teams in government to do exactly what the volunteers or the fellows or the brigades. So the question is what the great attitude of the IT services are towards the kind of intervention. So I'm visiting a scholar here and doing a project for the European Commission on co-production. So you are fully in the middle of what I do and I thank Code for America for everything I work from all your work. Thank you. So the question was what prevents existing IT departments from thinking about this stuff. I think there's a couple of different things. So we've seen a lot of different patterns in the different IT departments that we've come into contact with. Probably the biggest problem is a kind of jurisdictional question about who is supposed to be dealing with technology. So a lot of times you know cities will sort of assume that digital equals computers and the IT department is the computer department. So therefore digital dot dot dot like IT department. One of the things that we're finding is that that equivalence is really false for a lot of places. You know digital communications is a thing that's more connected to like interactions with humans and getting like kind of at scale people talking to one another. Then it is around like I don't know like whether the email servers are working or whether the web server is up or whether the the police radios work. And so I think that we've kind of seen this pinch on the IT departments where on the one hand there's much higher levels of expectation around like just the basic operational like functioning of the stuff that they steward like police radios and like email servers. And then from the other side they're all getting pinched really really hard budgetarily. So you know they have a lower budget to work with Oracle is more than happy to take their money. So they they're just like shoveling money out the door in the form of these like you know I think 10 million dollar Oracle contracts like the Oakland IT department has right now. And so there's not a lot of room in between those two pressures for somebody to think about new ways of doing stuff like they're struggling to hold on to just like run the things that they've been tasked with running. And at the same time they're having their resources stripped away from them. So one of the strategies that we've been taking is trying to figure out you know given that digital doesn't necessarily mean the computer people. How do you get digital stuff to happen outside of IT departments. And a lot of cases it's kind of teasing IT departments away from other departments. So like letting them get back to like email and police radios and making it possible for somebody who's in like I don't know Parks and Rec or communications and press really relations or something like that to be able to make arguments about their digital technology and work with the fellows to kind of introduce the human needs that they have in order to produce this technology. So it's really about kind of getting the IT department kind of out of the picture because they're just not the right department to be thinking about this stuff. I think historically IT comes out of what was the term like office automation. So it's like literally the IT department is the department that like buys phones and you know sells you a compact computer and replaces your laptop when it goes bad. Like those are not the folks that should be thinking about digital maps or communications or feedback. Is there any issue or communication between the business units in my place and I was a director general for IT and I have 1,500 people. For me the most important people were the development guys. The rest I have to do I mean it's house TV. Yeah. The lowest possible maximum reliability. Yeah. Yeah. And citizens don't always they don't always have access to like developers in these departments either. The business department that they think about that I saw in England some some people from the business department initiated projects. Is it because the IT department didn't have the resources of who they care is or and the second thing is once you develop something you have to sustain it. Right. So I mean how do you sustain with volunteers of that. Right. Yeah. So the question with the second question was how do you sustain the stuff with volunteers. That's a great question. I think you know one of the things that we've seen especially with like cloud based deployment platforms is that it's started to become easier to think about sustaining a technology than it used to be. So you used to have questions like I don't know hard drive crashes magnetic media going bad computers going bad having to replace them doing all that kind of operational level stuff. One of the things that we're seeing is that there's actually really nice kind of intersection of opportunity where on the one hand all the operational stuff is becoming easier because you have companies like Heroku or Amazon or I don't know Red Hat and others that are producing all of this stuff in a cloud based form where you don't actually have to worry about the hardware under the hood. And then at the same time actual city requirements for this technology are like cartoonishly tiny. You know they think that they have a really big technology problem but it's you know I mean Oakland City Hall gets like I don't know a few dozen website visitors at a time like we're not talking about like huge numbers of people. It's not Facebook or Twitter. It's just like you know it's Pawnee or something like it's small. And so I think one of the things we're seeing is that the ability to keep these things up is much less of like a intense operational need and much more of a long term sort of expectation meeting kind of need. So it's simultaneously easier and more accessible to people to think about that. Yeah. Indirectly. Yeah. So you're probably thinking of US Digital Service which is Mikey Dickerson Haley Van Wick and a few other folks that worked on the healthcare.gov rescue basically have set up a sort of executive level group to do kind of oversight of other I.T. projects. One of the people that was involved in setting that up was Jennifer Polka who's our founder. She basically departed from CFA for about a year between 2013 and 2014 to work as deputy CTO and the federal government helping to set this stuff up. Really low level stuff to set this set this kind of thing up as well. Like you know the questions that a federal government has to ask itself around this are not like you know can we possibly do this. There are more along the lines of is it legal for us to do this. Like are we allowed to do agile software development. Do I don't know does the office of the CIO have the power to tell another department that it can or can't do a certain kind of development. And I think once they started rooting around in the law around a law around all of this stuff they basically realized that they actually have much more power than they thought they did. You know they actually have the ability to you know specify how technology problems get done and specify how things can happen and hire people without having to go through a really crazy hiring process. So yeah USDS is probably one of the I don't know bright lights at the federal level. They're just getting started. I think they have like maybe half a dozen people working there right now. What's that 10. Thank you. Yeah 10 OK. So they're almost that a dozen people. But they're kind of slowly building this up and they have like really high level people there. There's you know Mikey Dickerson who led the healthcare.gov rescue Dana Chisnell is incredible. She's been doing like a decades worth of research in terms of voting patterns and how people interact with election process. She comes from a kind of ethnography user experience kind of place. So they're getting people from a lot of variety of background to focus on these digital problems. Yeah. I am so glad you're here. I really want to ask you this question. OK OK so you're somewhat associated with the election or the five year raw data and your project that I think was under Code for America. You're talking about census reporter census reporter. Yeah not Code for America. Oh it's not Code for it's not Code for America. All right. But we're friends with those guys. But sure. I mean yeah. So. Huh. Not tons. I mean so the question was about census reporter which is available I think at census reporter.org. It's a it's a project that was funded mostly by I believe the Knight Foundation and some other kind of news gathering organizations. It's an attempt to take U.S. census data and place it into a form that's more directly consumable by sort of newsroom levels of development talent. So people that are more interested in the story necessarily than the kind of you know large scale database thing that actually goes into getting the stuff out. I can't say much about it because I'm kind of merely friends with them and they don't really have a formal relationship with us but my understanding is that they reach the end of their grants and sort of produce the website and now you can have an API into this data. Okay I think that's more of a coincidence than anything else. Yeah. It was a two year project. Yeah. And then OpenStreetMap Legality of OpenStreetMap. It's still legal. For those of you not familiar with OpenStreetMap the question is can I say anything about the legality of using it. It's a crowdsourced data set. Actually I mentioned it way at the beginning of the talk because we made that watercolor map out of it. Has a license called the ODBL open database license and works by kind of consuming a lot of information from both government and non-government sources. So government in the sense of like pulling data from census. Non-government in the sense of like enthusiastic bicyclists with GPS units riding around and like marking down where all the post boxes are. Most of the questions about the suitability of its use tend to come from what's called like the viral clause in the license which basically means that when you consume this data you have to you're under obligation to kind of republish it the same way that you found it. And I think a lot of folks more in the kind of business world who are worried about the intermingling of their data with this stuff find it to be troublesome. But it's not super related to government stuff that we find we think of government as more of an input to open street map than an output from it. And a lot of the challenges that we've seen in terms of how government uses open street map data connect more to like levels of trust that government has in crowd sourcing operations to begin with. Basically if it's some dude on a bike who's like recording where all the post boxes are can we trust it can we actually consume this thing and connect something to it. I don't know that's a super helpful answer I don't feel like I have kind of the latest information on there. Okay I think you had a question earlier. Yeah very much so so the question was about the kind of trade off between if I understand this correctly like kind of going out for feedback from your citizens compared to the natural biases that you might just be hitting all the like rich people with computers essentially. We did a research project over the past year or so in Oakland where we really kind of Oakland is sort of like I don't know it's our spirit animal of the city or something like we're all from there so it's nice to work with Oakland. We did a bunch of digital communications research with Oakland where we found very early on that our responses were like disproportionately skewed north and uphill essentially so like where the white people live who are rich and have computers and like nice phones. Because we actually had a sort of census based way of determining that like because we could look at the map and say oh crap we're only getting people from here and not people from here it led us to kind of change in a sort of agile way what our research methods were in the middle of the process. So we were looking for information just like how people use digital services and really wanted to get as much of a kind of even cross section of the city as we could to understand things like mobile phone usage whether people had computers at home what languages they spoke that kind of stuff. And seeing some of the early returns on the stuff and mapping them led us to a couple of different sort of outcomes one was you know we immediately started to look at things like street teams and kind of more like flyer based distribution of information about this to try to figure out how to hit people in the neighborhoods we were missing but then the other half of it was literally just working with the communications department to show this result to them. Like one of the things we found is that a communications department in a city like Oakland is traditionally much more tuned to the communications needs of the press. So they know you know every single reporter's phone number and they know exactly what's kind of like in the wind in terms of political opinion but they don't actually understand that they have no ability to reach I don't know the Hispanic population living in the Fruitvale neighborhood for example and so where we left that was kind of in two places one was we tried to make it a little bit more equitable as best as we could given the sort of limited time and the other was we tried to leave them with a methodology for understanding how they could bring those numbers more in the line. So it's sort of like we're still in the kind of early phase of the feedback loop which is like well we haven't fixed it but we have at least identified and quantified a problem so you can kind of show this to somebody who's doing research and say okay there's this thing called the census and it tracks like what kinds of people live where and if you ask them about their neighborhood or the zip code or something you can actually get some sort of useful feedback there in terms of whether you're hitting like you know only the white English speakers or if you've actually got like the Vietnamese population or the Hispanic population or whatever. Yeah well that's not working. The acknowledgement that between the sewers category tends to be done with the punts most open source is something that I like and that I enjoy and that is I'm going to go clean the sewers and these are kind of the digital sewers of what you're talking about things. Yeah so how do we balance the divide between kind of the innovative fun stuff and the clean the sewers stuff. I think one of the ways in which we balance it is that we've discovered over time that I don't know one man's sewer is another man's innovation I guess. So we have a lot of folks in this kind of brigade community that get really excited about things like bus arrival times or you know open data or that kind of stuff. For us it's a really kind of tricky like edge to walk on because one of the things we've found over the years is that if you work with the political side like mayors and elected officials they go straight for innovation every single time because they're like they want to be seen as innovative they want to be seen as flashy. They will move heaven and earth for you instantaneously. The problem is that it's a very temporary sort of feeling move like a mayor will get a thing to happen in like the drop of a hat but then things kind of flip back to where they were as soon as the mayor stops looking at it. When you look at the administrative side so that's like non elected officials you know the IT departments the city managers that kind of stuff it takes forever to get them to do anything but once it's done it stays done. So we kind of have to walk between those two and it's almost like you're kind of like playing them off against each other. It's crazy I think our very first year we really fell into this trap like the first time we ever sent a fellow any place before we really knew what we were doing we learned really importantly that difference when I think our fellows showed up in Washington DC and were immediately ejected from city hall because an administration change had happened and it was the previous guy who said that they become and the new guy was like yeah no sorry so we've learned over time that you have to kind of work with both sides and we've really found that a lot of the people especially in the brigades are really quite animated about the clean sewer stuff like for them it's it's kind of a workability operational sort of thing like that's where a lot of the kind of meaty chewy interesting sort of data and technology problems are and a lot of times what they're looking for is just that like that little bit of recognition that says okay a difference was actually made here I think there was a situation just a couple of weeks ago during one of our national code across events where somebody from Baltimore helped the transit agency there map its data in a particular way and the headline was that it had saved the city like $600,000 because some vendor had quoted the city that price to do this work and it turned out to be like a weekend's worth of kind of hacky data work I don't think that's the thing that can necessarily scale to everywhere but I mean it sort of just sort of shows that a lot of these things that look expensive the cities are actually kind of quite easy for somebody who's animated and motivated by it to mess with it and a lot of those people are excited by the sewers yes the question was typical life cycle of the project and second question was civic commons what happened life cycle of a fellowship project is typically about I think 11 months is the official term and the fellows are all kind of in sync with one another so they all show up in our office you know the first weekday of the year basically we spend about one month you know going through extensive training so the people that join our fellowship are typically folks who are sort of mid career they might be like in their late 20s early 30s they've got about five to 10 years of experience of some sort under their belt so they already know how to code or they're already designers or they're already accomplished researchers and what we're training them on is not so much how do they do their job but what does their job look like in the city environment because they're coming from places like Google or other companies where you know they're accustomed to things working at a sort of private industry kind of speed and they need to start to understand what it looks like to kind of you know both work at the slower pace of government but also do all of that stuff that you have to do in a government like you know kind of gain people's trust and work with stakeholders and get them to actually care about what you care about and kind of give up information when you would like it in February which is ending right now they get sent off to their various cities so all of our fellows this year went to go live in their various cities for the past month in some reveal they had like 50% snow days I think so it actually was quite difficult for them you know folks went off to West Sack went to Vallejo went to Richmond went all over the place the purpose of February is to gain like an extremely detailed view of the actual situation on the ground in that city so they meet the people who are experiencing the problem that they're there to work on firsthand they meet local residents they interview people they engage in a kind of design and research process and the idea is that we basically send them into the situation with a kind of broad area to work in like do something with I don't know health care or this city is concerned about these vacant properties or this city wants somebody to think about economic development and the job in February is to figure out what actually is doable there and then the rest of the year is really freeform it's basically like you know the only other kind of punctuation during the year is our annual summit event which happens at the end of the year where people sort of present progress but in the meantime they're trying to figure out what it is that they have to do maybe it's data that's hard to come by that they have to spend six months getting out of the government or it's research with citizens and I have to figure out who all the local groups are or it's some kind of design process where the city is interested in like a new way of communicating with their citizens so at that point it becomes really freeform but the ultimate goal of the fellowship is to get sort of like alpha quality work out of them so it's something that demonstrates a solution demonstrates potential is useful for somebody in that one place not necessarily like a product or service that can be reapplied without any energy or work elsewhere but at least demonstrates that the concept is valid in that one spot. Does that kind of answer the question? Okay and the other question was about civic commons which is a website that we ran for a long time which was sort of a compendium of all these different projects and I think we've pretty much stopped running it at this point because one of the things that we found was that the demand for like a catalog of all these like civic hacks and civic tech projects doesn't actually seem to translate into like the use of them so we found that there was all kinds of stuff that was going up in the civic commons website that was like you know basically a catalog of all these works that were happening but because they were in the sort of typical alpha state like the result of a hackathon or the result of somebody really going deep on one particular place for one particular year they didn't have the same kind of reapplicability that we've seen previously or that we would like to see so what we've done instead of civic commons is we actually have a director of product now where product is sort of it doesn't mean sold things it means taking an alpha quality thing and making it a beta quality thing so taking like let's say city voice that helped that particular city in Indiana and then generalizing it somewhat so that other cities or other researchers could make use of it taking soft story and seeing if there's other ways to represent that data and other types of places or other data sets in the same place so the product initiative is really about realizing that you can't just kind of catalog all this stuff and have it be useful you actually have to like do things to it in order for it to be useful to people yes yeah so how would we work with non-digital versions of data I would say that in terms of the fellowship like when we accept a partner government to work with we are looking for sort of a baseline of some sort so if their IT system is like some water logged like filing cabinet someplace then there might not be something for the fellows to do there however we do have kind of partner companies that we work with I think one of the most notable ones actually came out of Berkeley a couple of years ago called the Capricity started by somebody here in the computer science department which explicitly aims to take paper forms paper information and digitize it so you know the fellows maybe can't help with a kind of purely paper based kind of situation but that there's different companies and other people that we see kind of starting to tackle these problems and kind of bringing things up digitally I think I was talking to a friend of mine who works at consumer finance protection bureau right now and he's really interested in these kind of parcel data sets and what would it look like to have like a national registry of parcels and addresses and his estimation was that you know the 3,500 or so counties in the US could kind of be broken into three groups like one group is they're digital they're online they've got stuff published it's you know it's you're good the middle group is they're digital but they're not online because they maybe don't want to publish this stuff so they have like GIS for their own internal purposes or they have digital services for their own internal purposes but they're not interested in releasing it perhaps because they think that that's too much of a cost or it's a pain and then there's this third group where he's just like I don't know they're rural like nobody's ever put their stuff into a computer before it's like a pile of paper at the county courthouse somewhere and I think it's that it's that third group that you know we're not even sure how we would approach them but it sort of feels like you know give them 10 or 20 years and everybody will sort of eventually hit digital at some point yeah what can you say about open addresses open addresses yeah that's actually a really fun project that I've been spending a ton of my time on over the past couple of months I didn't spend it yeah thanks open addresses is a project started by E.N.D.S actually who worked on the census reporter thing really cool project the sort of founding concept of open addresses was the realization that address information meaning like points saying where you know one, two, three main street was located in the world was not really appropriate for open street map both because of the license of open street map and the way in which address data needs to be used basically it's it needs to be mixed with other data sets where something like open street map maybe doesn't need to be so E.N.D.S who was working on the census reporter project also created a project called open addresses it's available github.com slash open addresses or open addresses.io if you want to participate and it's sort of a slightly different concept from open street map open street map is really based around kind of citizen collection of data and it's useful to like take you know two streets and put them in there because somebody will come after you and take the next two streets open addresses is based more on finding officially published sources of address info so this might be in the form of like I don't know zip files that contain every tax assessor parcel in Alameda County or point files showing every single address in the entire Netherlands or something like that and doing a kind of regular collection cycle on these official data sets with the aim of producing essentially a worldwide address database that contains points for like everything everywhere globally I think we're up to about 110 million addresses right now and it's you know if you look at the map it's spotty like we've got like I don't know all of Japan except for one prefecture all of Spain so that's great and then the US is kind of a patchwork like Arkansas has all of its data online California counties are kind of hit or miss depending on their digital maturity and it's sort of a patchwork beyond that is that helpful? cool yeah I was wondering if you can tell us the story of sort of long term impact trying to issue so you showed us these fabulous websites but you know now what like you know and you touched on this just a little earlier with the other questions that were saying what are the impacts you're having as we look like you did with that but what's your plan to do more than just have a fellow do something for a year and then it goes away? yeah so the question was how do you make it so that this stuff actually has kind of a long term impact rather than just sort of you know stuff gets changed for 11 months then it all kind of evaporates after that I think for the first couple of years this is still active I think we're starting to change kind of our belief and how this stuff works but for the first few years we found that just showing what's possible was really the core of it so I think you asked the question about blight status in New Orleans part of the work there was actually getting this data online into a form the citizens could see the other part of the work was getting this data online into a form the mayor could see and using a process the mayor could see so the mayor sort of had it in his mind that this was going to be like a six million dollar ten million dollar whatever you know six months twelve month eighteen month kind of process and then these goofballs from California come and do it in three weeks suddenly it's like you kind of have to recalibrate everything you know about IT infrastructure and operations and procurement in order to kind of accept that and I think a lot of the changes aren't so much that like we build a thing which then changes the world but it's like we build a thing and then people who whose actual job that is to change the world see that they have another option available to them probably the best example that we've seen of this is actually in the UK right now there's a government agency called the government digital service which is under the cabinet office so basically reports up to the prime minister there and their entire job is to be kind of like IT ass kickers so they go into these other departments who are spending you know sixty million pounds a year on some website do an analysis of what the thing is actually doing and realize that it's like a spreadsheet with some HTML in front of it and so a lot of it is sort of like looking at these problems that have been allowed to kind of grow into these sort of procurement monsters and figure out if there's some way you can strip the complexity out of there and lay bare the fact that it's just you know two lists of things that have to be like cross-referenced or a map that needs to have some stuff put on it or what have you so a lot of it is really kind of focused on that it's like how do you how do you convince somebody that their problems aren't massive that their people problems aren't technology problems I said that kind of our view on this stuff is sort of slowly evolving over time I like to say that we're going from showing what's possible to doing what's necessary so what we found over the years this is kind of the genesis of a lot of the work that we've been doing in Oakland but you know when we first started doing this people would say oh you do websites that's so cool and we'd say no no no we don't do websites like that's not really our thing and so we spent you know all these years we kind of ignored it laughed at it like fought it and then it won and like so over the course of the past year we basically realized that like the humble city website or the humble county website is actually like truly like the one digital thing that every single entity of any kind in the world including governments actually needs to have it's the one thing that everybody's guaranteed to spend money on it's the one thing everybody is really kind of tense about and it's also the one thing where like things like you know mobile usage of technology and so forth is starting to really kind of hit these people pretty hard they're really spooked about mobile in particular so we're doing a lot of work around website publishing specifically under a banner that we're calling digital front door to address the idea that there's actually some pretty mundane looking things that are the locus of a lot of really kind of weird technology decisions and a lot of weird problems that they're having the metaphor of the digital front door is you kind of have to have exactly one so you actually end up with like some interesting questions about replacing existing infrastructure you know packing light as you move data from like one form to another doing all the resident research that allows you to you know determine what languages you use and what what kind of terminology you use when you talk to people so it's this kind of move from like you know just sort of spinning plates and doing magic to like kind of getting into like some infrastructural type stuff yeah thank you so my question is is that a country? mm-hmm did you mean like staff or let's say the volunteer base? ah both both so the volunteer yeah the staff and volunteer ratio is huge I mean I think we have something like 40 staff probably and then the volunteer base is like 18,000 last I checked and that's spread all around the world all around the country you know it counts people that show up like once a month at meetings so it's sort of you know it's people that are like interested enough to come back periodically they come from all over the place you know when we've talked to them it's been people who were design professionals technology professionals you know in some cases just kind of like local gadflies that want to make a nuisance of themselves we kind of get all kinds and it's all just wonderful within Code for America in terms of staff some rough numbers my team technology team is like a half dozen people and so that includes designers it includes user experience people and it includes programmers the organization as a whole also you know because we're a nonprofit we have revenue we have people to talk to funders we have a very strong community wing so there's a whole group of people whose job that is to do all the kind of community organizing around things like Brigade there's probably about a half dozen people there as well led by Catherine Bracey and their whole deal is you know they are technologists but their background is much more supportive of the idea of you know holding these meetings getting people kind of on the same page holding synchronized national events flying around giving talks that kind of stuff trying to think of what else yeah and then just basic kind of operational marketing kind of stuff we have a whole team of people that leads what we call the focus areas so we have three of them it's basically a safety justice which is sort of like both public safety and criminal justice all kind of in one thing so Jennifer Montia Tanzi runs that she's like you know PhD and something related to that field working there Rebecca Colias is our health lead so she's the person who oversees the folks that are working on things like Connect and Balance and you know these food stamp programs so it's health kind of broadly interpreted to cover public health and things like that and then Alicia Ruo is one of our former fellows from a number of years ago who leads our economic development focus area so the work of economic development is things like opportunity permitting business process all the kinds of stuff that like small local businesses would need from their local government she's there to kind of sew all of that work together and make sure that we're actually doing it in a way that is consistent with how you should that's a one word question okay yes hmm yeah hugely desperate so we've been using GitHub probably since 2010 so the 500 projects are basically it's literally everything we've ever done typically there's only like a small slice of that that's being worked on actively so you'll find stuff in there that ranges from like long running projects that have a lot of contributors that are actually very active all the way to somebody's weekend hack for a hackathon someplace which then got kind of abandoned our theory is that you know because GitHub shows stuff in most recent first that things that kind of bubble to the surface will be up front and people who care about that stuff will be up front and then everything that's kind of you know smaller older not so cared for will just kind of sink down to the ocean depths at some point we do however have something called the CF API if you go to codeforamerica.org slash API you'll see some documentation about it where we're looking at a kind of we're trying to sort of holistically look at that entire structure and attempt to measure things like which repositories are active which things outside of our organization so like you know partner organizations you know Brigades sometimes will have their own GitHub repositories that kind of thing what's being done outside and then we're also looking for some specific things like who's working on stuff who's committing stuff and I think most importantly who has flagged an issue with the phrase help wanted so if you go to our homepage codeforamerica.org and scroll down you can actually see something called the civic tech issue finder which is a selection of issues from that API so our GitHub and other people's GitHub accounts that people have flagged with help wanted so if you want to code if you're interested in helping something basically go to our homepage scroll down and you can find some information about what's currently ongoing and what you might actually be able to help with does that help okay cool I think that's it alright thank you so much cool thank you cheers