 It's time to get started. I'm Cliff Lynch I'm the director of the Coalition for networked information, and I'll be introducing this session very briefly. So, you've joined us for one of the project briefing sessions for the start of week two of the CNI fall 2020 virtual member meeting. And it is focused on organizational and professional kinds of issues. And we have a really timely presentation here. A couple of mechanical things we're recording this session. There is, and it will be subsequently available. There is closed captioning if you want it. There is a chat box and feel free to use the chat. You can use yourself to share thoughts as we go along. There's also a Q&A box you can use that to pose questions for our speakers at any point. What we're going to do is have two presentations and we're going to take a Lauren lead off. We'll take a very short pause between the two and field any questions that there are before moving on to Scott's. And that's about and those questions sessions will be moderated by Diane Goldenberg Hart from CNI. I think that's about all of the mechanics that I need to touch on. We have two speakers today, Lauren Michael, who will lead off she's from the University of Wisconsin at Madison, and Scott Yockel, who is from Harvard University will finish up the presentation. Just briefly, this is a topic that is very near and dear to my heart. Those of you who are CNI regulars will undoubtedly have heard me say more than once that one of our great challenges at present is to do what we spent the last quarter of a century doing with instructional technology to professionalize it and find a place for it inside our institutions. And the challenge now is to do much the same thing with research support research computing and data. I'm delighted to be able to bring you today. Some of the folks who are involved with an organization called CARK that you're going to hear more about in a minute. That are on the front lines of leading this move towards professionalization and institutionalization of these critical roles. I just finally want to note that this presentation and the presentation that took place just before it at two o'clock on benchmarking research computing and data capabilities. I believe form a very good pair of sessions so if you didn't see the one at two, I'd urge you to go back and look at the video of that as your time permits. And with all of that lead up. Thank you so much, Lauren and Scott for being here. Over to you, Lauren. Thank you, Cliff. That was a great introduction. And as Cliff mentioned, I'm, I'm Lauren I'm at UW Madison and beyond the roles that keep me busy generally at the UW and also with the open science grid. I also have various roles within CARK, which is a research computing and data community that we hope to speak with you about today as well as diving into some efforts within that organization to support professional development and to help institutionalize and formalize research computing and data roles as a profession. So Scott will take that second part. I'll first discuss what CARK is how it formed and how we support professional development through the people network. So CARK, which stands for Campus Research Computing Consortium is grassroots organization of professionals that operate advocate for and advance the state of campus research cyber infrastructure, and also do the same for associated professions relevant to campus cyber infrastructure and if you're not as familiar with the term cyber infrastructure. So I think research computing and data, which is really the terminology that we tend to use on the CARK website, where we mean to be fairly inclusive of anybody who supports operates, not just computing and data infrastructure, but human support for researchers who take advantage of that infrastructure, whether that be computational clusters, data storage and creation services, even really data scientists and people who work with software. That works on data for research and CARK kind of pulling together from a history of collaborations and projects among this community of individuals, some of them leaders of of campus services. It really does sort of several things. One, it provides a professional community, whereby individuals who work in research cyber infrastructure can learn from each other. In what we call the people network that I'll describe in more detail that we think of as sort of a virtual ongoing conference. We also provide infrastructure like email lists, Google groups similar to what we use in the people network to organize members of the community into working groups that that have been identified with specific objectives or products that could benefit the greater community. What you'll hear about later from Scott is some of our work on professionalization and workforce development for roles relevant to research computing and data. In the previous session, Cliff already mentioned that there was a presentation on outcomes of our working group on developing a capabilities model for institutions to assess the sort of state of computing and data services and capabilities on their campus. We have other working groups but another one that has had sort of frequent reinstatements of sort of just surveying the research IT and research cyber infrastructure structure ecosystem is a working group that has executed some of the CARK sponsored workshops, surveys of research cyber infrastructure services, and there was a paper submitted to the PERC conference practice and experience and advanced research computing that had a snapshot of that ecosystem that we have other working groups that focus on things like the CARK charter, etc. And then we also support the formation of interest groups that might lead to objectives of working groups but can otherwise just be a way for individuals to discuss issues relevant to the ecosystem. In this diagram here that we sort of demonstrated the set of activities, we have a logistics team that really is just in charge of maintaining some of that common infrastructure like the Google groups and email lists that support the people network and these working and interest groups. And I'll talk a bit more about the people network now, which includes several tracks that professionals in this field might ascribe to and therefore engage with. Scott, can we go back to the principles and charter. Yes, so a few things that we like to communicate about CARK that are really sort of built into the principles of how CARK operates and how it supports community activities. We try to be very inclusive and open we think of all of these activities as community owned owned by the community of professionals themselves and not say by CARK. Therefore they're open there for the products that that we put out can be used and taken up by anybody. We try to facilitate and span boundaries there are a number of other organizations and the research computing and data space that either represent professional development and community building opportunities. So we tend to partner with those and try to make sure that we're representing everybody who who touches or thinks of themselves as being involved in research computing and data. Next slide please. And so in our work with the people network, you know we started by looking across you know some existing communities of professionals but that were maybe subsets of the research computing and data community. And we were thinking you know if you needed when we need to pursue something that's relatively nor cutting edge, where do these individuals turn for professional development, a downside currently. I think for this field is that there are some formalized trainings etc that that tend to be more oriented toward enterprise IT and that don't necessarily target research cyber infrastructure specifically. And so what a lot of us do is that we talk to our friends we go to conferences and really in reality we do a lot by doing we learn a lot by doing and and by doing research even in cyber infrastructure alongside running cyber infrastructure that supports research. But conferences are kind of generally only once per year. And there certainly has been quite a bit of success in sort of email centered or call centered communities that we thought we might bring to this inclusive community. And so, next slide. So we started the people network representing different sort of facets or what we've called facings, which Scott will elaborate on, and this is how you know one of the frameworks that prior work and sort of workshops by members of this community have, have kind of put together to express the different roles of the profession so there are individuals who are researcher facing meaning they do consulting facilitation outreach and training. Some people who run computing and data systems, including hardware network components, people who are data facing, including librarians people doing analysis preservation. There are leaders. There's a software facing entity within these professional roles as well. We haven't executed that track yet within the people network partially because there sort of already isn't a community that we collaborate with called the research software engineers community or RSEs that you may have heard of. So next slide. There's a date where each of these communities has an email list monthly calls and a couple of coordinators who don't really lead or own but just help each track or sub community organize themselves. This is the sort of breakdown of membership so we've got hundreds of institutions participating. We have a number from edu. Of course we have some email addresses that are our Gmail but we're sort of trying to represent with email domains, maybe something close to the number of organizations with professionals participating. And we've had quite a few people go out of their way to sign in to call documents. We also record a significant number nearly all of these monthly calls. We have a number of them on YouTube so there are quite a few views there and and zoom connections as well. And next slide. Yep, this one. I've gathered if you're interested in perhaps joining the people network topics that have come up within each community that they've decided they wanted to discuss, or that one of their community members thought they could present their own story on. The format for these calls is not always like a one speaker presentation sometimes it's a group discussion with absolutely no slides, and really the each track decides what topics they want to discuss, and that sort of thing and then there's also quite a bit of email chatter before and after these discussions. The main people network lists that's inclusive of all of these track lists is where a lot of people also post job positions. So if you're looking to promote those or to be learning about positions the people network can can help you just with that as well. We've got a number of calls coming up including an end of your party in December, where most of the tracks are not having calls in December given conflicts with the winter holidays and you can learn about those on our events page. So at this point we'll pause just briefly to take maybe a couple of questions about the people network before Scott discusses a bit more about professional development or development of the profession I mean. Thanks, Lauren, if you have questions, please enter them in the q amp a box. I just quickly wanted to ask Lauren, is there, are there opportunities for students to interface with professionals through kark. Currently we don't have anything that would restrict students from participating in the people network. And, and we certainly even have discussed in some of the communities I think one of the topics actually was student employees. So we do have students who are engaging. I think that's an area where we could engage more. Okay, thank you. Okay I'm not saying any questions come in through the q amp a right now. So thank you for taking that break. Yep, we can do more at the end. Thanks, Lauren. So I'm Scott Yoko Harvard University. I've engaged with kark, I guess since the beginning, and Lauren and I and a few others across a couple of institutions were part of the ACI ref program, which was the NSF grant called advanced cyber infrastructure research education and facilitation. So a lot of the like understanding of how to operate as a virtual organization started out of that kind of space and then kark formed shortly at the end of that grant. And it's kind of interesting now that the norm is to have these conferences and these conversations all over zoom. We were all really well practiced at it and had all these skills that we already built from that. And so one of the things that we all struggled with and it became interesting at that point of the end of the ACI ref program was that the concept of a facilitator as a position was actually relatively relatively new, and we were starting to see whole organizations like create and want to have something like that. So, we started to talk about why having professional like why we need to professionalize and think more about research and computing and data. So what things are, there's a national shortage for these people. There's national shortage for it in general, but there's especially a shortage, most of us go through long searches to find people to fill these positions. There's little awareness that this is even a career. There's so many postdocs and graduate students that that are in the modeling and simulation space that don't realize they could continue to stay in academia and transition into this, this kind of support and research, and not, you know, not choose other paths. There's also a high employee turnover which is hard. We're in the business of educating people in university we do it also within our, our staffing because we are doing something relatively new. And so those people, like in the Boston area, my, my employees don't have to move anywhere to take a job at Google, or, you know, or any of those types of places they can just go down the street in Cambridge. It's also hard because it's hard to train them as we talked a little bit about that before. In our structure at our universities developing and promoting them as a challenge because there are not. There's just two classes of position like an entry level class and then the senior position, anything beyond that was management, which is not always the best path forward for some people. It's a distinctive occupation from administrative or enterprise it in the fact that those those positions tend to have a vertical depth and less horizontal breath to them. And they tend to create things that are more of a utility for the canvas instead of trying to craft solutions alongside of the customer, like the end user so for us for researchers. And there's a lot of times when the things that we are trying to deploy are unique. They're not, there's no manual on how to install that software there's no, there's not very much knowledge historically about how to run that you know file system or something that's fairly new. So, that led us to do a lot more. So there was a series. So I'm walking through a little bit history of a couple of years there was a series of meetings that was part of a research coordination network that started I think 2017. And a group of us about 30 or so director of research computing and a bunch of organizational sociologists started to talk about what does professionalization mean. So in the library since you've gone through the very formal sense of it you got all the way down these steps right to create formal graduate programs in schools of library information science and stuff like that. So we started with research computing or we might say research it as you'll see a couple different kinds will change the terms. It looks like probably steps 123 would be nice as a place to sit to hit at this point where we get to the point where we, we organize, what is the knowledge base, we disseminate it as a practice. So we started to organize who the people are and organize like what management looks like what individual contributors look like and stuff like that. To go beyond that would would make it a much more rigid kind of situation where we need something to be continually updating and changing every couple years to go along with the path of the research. So we think of ourselves as kind of stopping probably after step three there. A working group that's kind of keeps revolving every year and coming out with different products. That's a part of supporting professionalization. And the goals really is to develop and disseminate frameworks and approaches to use with institutional leaders in HR and it and research, so that we can elevate research computing data as a distinct and highly valued career path. So we look at the libraries and the libraries have done that historically and done a really good job. There's a, there's a, there's an elevation of a librarian at a lot of universities that is similar and has similar tracks as professorships which is, which kind of shows that that value. The value that we can also provide is attracting retaining diversifying the talent that we need to support research computing data, right and developing new practices in that space. In 2018 we got together and tried to think about these things and write up a document was kind of like a career guide about and define what are the roles and responsibilities of these jobs. So we got together and learned about what some of our colleagues had done thus far in that space. We developed a flexible framework and wrote about a 30 page paper on that internally. And we vetted the concept of these different facings. And they actually held that was one of the things that we're interested to see come out those. Those all solidified and everybody agreed we should stay with this. So then we wrote different sections defining what those were. The other thing is to be able to understand the way to communicate the differences in the facings and different levels of interest and needs of different staffing and understanding that we have lots of varieties of different institutions, small, large private public, you know, from Epscore states and from not, and we need to understand how to build a framework across that all of that together. The next step that came out of that was the result that we needed some also an actual framework that we could use with HR like the structures that are in place we need to be able to provide them those things and not come from the fact that we're just developing them individually at our own institution but it's something that's being used across all of the US academic space. On top of that, that workshop, some themes came out that have resonated and continue to resonate today as you'll see get a slide towards the end. The idea that we co create or partner with the researchers to come up with solutions. Career paths are important, and we don't have that most. The idea that there's an increasing amount of things that are just digital in nature now that is causing what used to be for us mostly high performance computing very centric to that. It just keeps growing and now we find ourselves overlapping and doing a lot of work and partnering across campus with data scientists and with digital librarians and other other communities. I talked about the status part as well already the value that the people the staff would like to have and be known for the terminology as you've seen in our slides. The term cyber infrastructure really comes from the National Science Foundation has an office of advanced cyber structure which supports a lot of our institutions as far as for grants. So, but we use the term research it we use the term research computing and data. It's hard when we keep using different terms to mean different things so it's important to understand and define these terminology. So, fast forward one year past that. So if you go to the products page you can see the 2018 document that we created in 2019 we set forth to create a working group to create an actual HR framework. What was nice is we had a good collection of participants in the working group. So we had some people that were directors we have people who are managers we have individual contributors we have people from HR that support the IT groups, both from Harvard and University of Florida. So this week's time, we went through this process where we come up with a skeleton draft and email it out on the call we review it and we just kept doing that and we go each week through each series, which I'll show you in a second, which are the facings to create actual job descriptions. I obviously can't show you some like what is 15 pages of an Excel sheet in a presentation, but I'll try to give you an overview of what a little bit of structure that I've learned about that happens in HR. So the top level the highest level is job function. So information technology research libraries normally is a job function. You might have museum or collections you could have athletics like those are the top level functions. And so at an institution research computing and data would normally fall either in the IT technology space or in purely research right. But underneath that select in this example information technology, there's IT infrastructure, there's networking, there's IT infrastructure say for cloud. But there is no IT infrastructure or positions there that describe really what happens in research computing and data professionals. What we really needed was a unique job family that was just research computing. And so that's what we set forth to create. And in that there are job series. And that's where the facings come in. And I'll break these down. So we just took systems facing the individual contributor roles. A lot of us just had two of these prior to this, this model. And in this model we thought it was good to have a systems professional one through five. Most of us are current staff fall into the category two and three in this space. And we don't have the way to just didn't have the way to describe a ladder going beyond what would be the individual contributor role. That was equivalent to a manager role. And so we were losing a lot of staff at that moment, because they would become technically more capable in the systems role. They would take on more responsibility. They would broaden their horizon for the number of types of things that they would become, you know, experts on, they wouldn't just only have to know about systems that also have to understand how researchers use the systems and have all this kind of stuff. So having this gives us the ability to have what we would consider like, if these were engineers, this would be at the level four, like a lead engineer or principal engineer principal architect. Those things also translate well to what happens in say like Google and Intel and these types of industries. They're in the technology space. So we have this one through five. These different facings. And then you can see how they're equivalent now to the manager. So this could be like a senior manager position junior senior manager or you might have an associate director just being on what but you can see there's a couple of levels. So within the position itself and in the framework itself. These are all the components that are requirements for for most every HR business title, the core duties and the additional qualification and skills are the parts that are. You have the most flexibility and then we encourage the individual institutions to adopt and change as they need. But the part that is really important that we try to keep static is the actual job title, which are just these that I showed before. And then the job summary in the way in which there are great levels. The thing that's the most important is the language that we use to describe how one goes from levels 123 and up up the ladder is really important. They're really important key terms that we learn from HR to keep in there to keep the letter going directly so if you go to the part products page you can go to the job handling matrix and see that information and see the full details of so I encourage you to go there. And a couple, couple more slides about what came out of that working group. I've already talked about the non managerial career paths, the distinctness from enterprise for it. One concept is that an enterprise it as you go up to the ladder, you typically become more specialized might become the, you know, architect of networking or the architect of Oracle financials or something like that for university. For research competing, you get much broader, you have to know and be able to traverse across lots of different disciplines, lots of different kinds of technology and lead in that space. Also, research competing is really just designed for those that are in that space. They're also going to be other people in your department that support you business operations like that. Those are in other families. And then there's a lot of argument about education versus experience and we really landed on the side of experience and that we describe the number of years of experience where education counts as some of those years of experience as the important thing that also keeps from alienating people from applying to positions where they think their education may not qualify In software and data, it turned out we decided to combine those for HR purposes. The roles and responsibilities of software and data professionals supporting research are really well aligned. It's just that their skill that they're using to do those roles and maybe the domain in which they're doing those is different. So it didn't you didn't really necessarily need a distinct separate series for those And one note is that when you implement something locally, you have to be flexible. So some of the hard things to decide on a where do you put data sciences. Are they research function or the technology function, you need to think about how that benefits them. Bioinformatics, as an example, I really encourage people to think do they is the primary responsibility of a person research independent research and collaborating with other researchers or is it providing a technology service. And that's kind of where you can make the distinction between the two. And at Harvard, it happened to be that the leadership facing types place is it's a whole other category doesn't fall into the framework. You will get pushed back when you implement something and you have individual contributors where there are very few or none. They don't want to implement those categories in HR. But if you don't have them, you don't have a career path for people internally within your institution. And then just recently in 2020 here just a couple months ago, we had 100 participants across three webinars where we talked about all the different types of things that we've talked about thus far. And a couple outcomes from that is that we really, really need HR job families, which was good that we already had done some of that work. The idea that we can define career arts that we need, you know, a good understanding of the organization for positions across the landscape of academia diversity equity inclusion like what is what is our current state of that how can we recruit and do better at that. And so one thing that's nice. Park, you know, completely not running that that workshop that was an NSF Thunder workshop that Purdue hosted. We already had in place some ideas to do this we are just starting a working group with an annual census for research computing and data professionals. Patrick who talked before us about the capabilities model career arts is a working group that he is getting started. We want to continue doing work of, you know, taking the best of training and workforce that exists in different institutions and kind of like centralized that information, especially working with like graduate students and undergraduates as they are the next kind of workforce into the space. And then other things like providing lean practice to support our city as an organization like how do we help support managers and how do we help grow their capabilities. Oh, that gives you a good overview of type of activities we've done to help professionalize research computing and data. Excited to answer any questions you have in that space and please check out the products page. You can see a number of different products we've created. Always welcome your feedback on those products. Thank you Scott. Interesting overview. Appreciate that. And if you have questions please go ahead and type those into the q amp a happy to take those now we are a little bit past time I apologize for that. But while we're waiting to see if we've got some questions I believe Cliff has a question. Yeah, I do actually and this is, this is a little bit off your primary topic but I would really welcome your views on this Lauren and Scott. C&I has done a lot of work over the past six months or so on issues around research continuity and research robustness, and a significant piece of that is about setting up experimental apparatus for remote operation and more automated operation so you don't need as much physical presence in labs. And that includes sort of core instrumentation that may be shared among lots of other projects. Where do the people who do that kind of work fit in. Now, I'm a little uncomfortable about thinking of them as research computing and data, but I don't know where else to put them either. Yeah, it sounds like you're talking about people who operate instrumentation. Well, design and interface and networks and things like that. Yeah, so we've actually carcass part of a discussion that came out of the NSF workshop that Scott mentioned. And, and I've been part of those discussions but there are several communities across research computing and data of professionals like research software engineers campus champions some of the other collaborators on one of my earlier slides. And we're getting together talk about potential community of these communities sort of one sort of unifying body, and we just discussed this issue of people who support instrumentation who touch quite a bit of computing and data but that may not be all of what they do. And I think similar to what Scott described for data scientists, there are perhaps different reasons at different facilities or different campuses where sometimes it will be advantageous or more logical. Just in terms of the work that a particular person does to think of them or as research cyber infrastructure and to that extent. I believe that a community of communities for research computing and data should include those individuals and should otherwise have the higher level objective of including emerging areas of research cyber infrastructure where I think automation like you mentioned and instruments and other cyber infrastructure tied to instrument facilities is extremely relevant. And one example that's probably done at the best is in bioinformatics right there are, we've carved off this community of people who are very technically minded but also it's still in that very much in that domain of research that bridge the space between an instrument that somebody who's a geneticist creates data, then that data has to be usable to make decisions and make discovery. What's hard is there are a lot of other places that, especially in microscopes that are like scan all sorts of things from materials to cells and stuff where we don't have that kind of framework. And, and I think also, as digitization spreads across like we see this at Harvard in the library system where they're they've automated the way to scan materials that have been handwritten, but they don't have a way to automate using machine learning and stuff to transcribe and digit like actually make actual text out of that person. And that's where there's a lot to be learned about where the problems are, and get the right people together, and you have to have kind of an end in thought about it like where is the data being created where is it being touched along the way. And where do we intervene on the other end on the systems and to make sure that we're providing things that work well, because if we just let the person who's creating the data, then do the whole pipeline. They don't have the visibility of how perhaps poorly they're using the computing and storage resources on the other end, right. And so they could do things in a much more performant way or do ways, you know, that would make more sense. So that is a challenge. That that's that's a very helpful perspective on that, which I've been puzzling over lately. Thank you so much. Back to you, Diane. All right. Thanks. Thanks so much. Lauren and Scott for fielding that question and Cliff, thanks for the question. I don't see any questions in the Q amp a and we are well past time we have another session coming up in about 20 minutes. So I will thank everyone for joining us and a special thanks to Lauren and Scott, and hope to see you at another CNI breakout during our fall 2020 meeting. Be well everyone take care. Bye bye. Many thanks. Thank you.