 Hi everybody! Oh, now I'm very loud. This is Elevating Cloud Native Education. So in other words, we're going to talk a little bit about how we feel like cloud native development and cloud kind of computing is not usually well represented, especially in an undergraduate degree in universities these days, and hopefully what we're trying to do to solve that problem. So we're actually going to lead the talk with Anuisha talking about what she walked into industry with, or lack thereof, and then kind of pair that with what we're trying to accomplish. So just by way of introduction, I'm Langdon White, and this is Anuisha. And in case you couldn't tell us apart, here's our pictures. But then I come from about 15 years of boutique software consulting, then about in nine years or so at Red Hat, and then I've spent the last two or three years at Boston University as a clinical professor, and Anuisha. Hi guys. So I'm pretty nervous. This is my first time talking at a conference. So I'm a grad student currently. I used to work at Oracle as a software developer for three years after my undergrad. And then currently I just finished my master's degree. I'm currently defending my thesis next month. And I'm also an incoming student to the PhD program. So my work, like a little about my research is that I work on systems, which means technically like networking and distributed systems, streaming systems. And my research area specifically is on database optimization. So I tune databases and I work on like auto tuning on specifically things called log-structured mergers, which is like super popular now. There's things like DynamoDB, RocksDB, which is what I concentrate on and LevelsDB, things like that. So I work with this really nice guy called Manos, who's pretty in the field of like database research. And that's a little about me. My experience in cloud is quite limited, but well, that's why I'm here. So I'm still like kind of struggling to learn cloud. I've just gotten into it. And then I'll just share like what I've gone through. Okay. So this is a little about like my experience at Oracle. I started out as a software developer, which if anyone else has started out as one, my role was not very clearly defined. I worked on something called Oracle Jet, which was our cool other language, which we were trying to put our whole UI, transform our whole UI into. And then slowly I got kind of tired of it. So I moved into more like in stuff, which was like the Oracle cloud when I was working for Oracle. So I worked for the Oracle Cloud infrastructure, but my product was a legacy product, which means that we couldn't build things again, like using microservices. We just wanted to deploy it on the cloud. So the two cloud infrastructures that we concentrated on was OCI, which was our own cloud, Oracle's cloud, and then Azure. And then what I was given, which maybe people can relate to, I was a fresher, so I didn't have any idea of what to do. I was given access and I was told like, go ahead and deploy things. This is your RPM package. This is your cloud. Go into it. So I went through the interface and I had no idea what a shape was, whether I needed backup, what was subnets. I didn't know how things communicated with each other. So I just kind of came back with questions. I was like, what is this? I've just run things on a VM. I don't know what a cloud VM is. So that's how I started out. And then I was finally able to deploy things on the cloud. And then I continued with the cloud. I ran several performance benchmarks because I used to work on OJET. We had this API speed testing and stuff. So I also worked a little on Microsoft Teams. But then I still had questions because when they teach you things in industry, they don't teach you why you do stuff. They teach you the steps of how you do it. So I just did that. I just knew that you had to click this button and just be like, oh, this is your region. And I just didn't know why I was doing things, basically. So I went back home. And I wanted to go into this other team, which was kind of the cool team, which was based on the microservice level of architecture. So I wanted to learn Kubernetes and microservices and Docker. So I went back. I did my research. I learned all about it. I learned the theory stuff of it. And then I came back and then I was told that you need actual practical experience to get into this team. And the nice part about that team was they got like MacBook Pros and I was given a ThinkPad. So I just wanted to go there because I was young and I just wanted cool things. So that was my primary motivation. I obviously didn't get into that team and I stayed on my team. But moving on. So which brings us to the issue that we have is that what often you learn from like textual reading or like videos is not really what you are going to be experiencing. So when I started out with like doing videos, I had no idea what issues you could run into because they gave me something which worked absolutely fine. So what I learned as a software developer was more often than not, things did not ever work out. So which brings us to the fact that you need more practical experience than theoretical experience. So this is my experience as a grad student. I tried getting into cloud. I've always tried getting into cloud because I thought it was new and cool and I could run things which I couldn't otherwise because I didn't have enough space infrastructure, I guess. So I took the streaming systems class because I was a very systems oriented person. I've always been into databases. And this is my professor. I was super like fascinated with her so I put her picture over there. If anyone knows her, she works on Flink and she's a contributor to the open source repository. She works on graphic, graphical, I think the library, whatever, whichever deals with graphs. And I worked on her on a project which was going to be done using AWS Lambda. And I was just given this project that like, if anyone's ever used like a streaming systems pipeline, which is like, you have this incoming traffic at sudden times and then your pipeline basically gets overloaded. So you have something called back pressure which makes everything much slower. So our idea was that when you have something which is causing issues in your pipeline, you want to use something which is readily available. So AWS Lambda was a serverless function and we just wanted to use that because you could just use it and then forget about it. For me, I was just getting started with Flink. So Flink was hard for me to understand and I had no idea what a serverless function meant. So I didn't know that you could just like call it using HTTP methods and then just like forget about it. I didn't know anything about it. So I was given this part of like invoking the Lambda function using my Flink pipeline and I got super stuck and that's when I started reading about like everything that AWS had to offer in terms of cloud. So the next thing I did was like an internship. I started working with Professor Langdon and that was primarily because like I wanted experiential learning and what Langdon works on sort of in our at BU is a sort of lab which tries to cover this difference between what you learn in class and what you want to learn at work. Like when you go to work you have certain things like people expect you to know certain things which they don't teach you in academia. So when I first started working they just like threw a bunch of stuff at me and expected me to know it the next day and I really wanted to help other people like get over that. So I joined this organization and Langdon actually has a really cool subject which is like we take projects from clients and we try to build solutions which are like built by students but the issue is here that the more freedom you give the more choices you have. So students had to decide their own text stack. They had to decide how they would deploy it. They had to make arguments for it and most people didn't know like undergrad at an undergrad level I had no idea what I was going to be using and I didn't know the pros and cons of that. So I joined this group to be able to work on this and then I started working on this project called SCAN. It's a data management system and what it does essentially is that if you have data available to you like publicly available data it helps you like host it store it and display it. It has its own like theming libraries and you can just use it out of the box. So a lot of government websites actually use it. In fact the Boston government also uses it and that's a bunch of like websites which use it and it's pretty cool if you guys want to do it. And then I had this experience and I thought it was pretty cool so I wanted to get into more like open source stuff and then I started working more and more with Spark and I also started working as a TA. Coming back to our discussion about the experiential learning lab the issue we faced was that students were not aware of like text stacks and things to use and how do you deploy apps? Like what is Cloudflare? What is Versa? Why should I use one over the other? If BU sponsors most of my things I can just use the most expensive thing ever. So that's why we introduced another course which would be a precursor which teaches you practical things more like why would you choose something else? Why would you use some other database? Why would you use Postgre over some other something? So that's why I became the TA for that course. I've been working as a TA for almost a year now and I work with Langdon a lot so I put a picture of that just so that you know that is associated with him. And that's our first class that we had that's a picture of all of them and that's the Spark logo just for reference. So that's like pretty much all for my part for like summing up exactly what I mean by sharing all of my experiences with you is that this has been my experience over like a very long period of time. So I did my undergrad, I finished it in 2019 I worked as a developer for three years a master's degree of two years and I'm still here and I don't think I can confidently say that I know everything about Cloud at all. So which brings us to the point that like when I talk to the people from industry at this conference there is a lot that I get to learn like from scratch because current academic syllabus does not ever cover what exactly you need to know about Cloud and there's so much about Cloud. There's like deploying on Cloud, there's building the Cloud and there's so many other aspects of Cloud. So we propose sort of practical learning as opposed to the theoretical learning that we've all come to know which is their syllabus and that's what Langdon is going to be talking about. So, oh okay I have another slide. So these are the things that I assumed would be taught in a Cloud native class because these are things that I had no idea about. One was Cloud deployment because I didn't know what deployments meant when I started out. One was containerization which is the most important part of the Cloud because I wanted to run my program anywhere I wanted to and orchestration tools which you cannot understand unless you know what containerization is. So I started figuring out what orchestration meant which is not something which you learn about either. So these are the three things I assume would be covered but I'll let Langdon speak for what exactly he's going to be doing. Thank you. So yeah my own experience with my students right is you know actually I have a class this semester for example that I was doing data engineering at scale and the first thing I did the first day of class was kind of give a Python quiz and I was terrified by the response on the quiz. You know so the kind of tactical hands on experience components you know what you're often taught in a lot of kind of CS programs in general is kind of like how to program right or how to find algorithms but not all the other things that surround kind of software engineering right or software development which is all those things like deployment and how do you you know like what is a container you know many of my students actually have never even used or seen a virtual machine they don't even know exactly what they are so you know like when I in almost all my classes when I want to explain containers I have to start with okay well back in the day we had these physical servers and this is the problems right and then I kind of lead into virtual machines then I lean into containers and then sometimes for the more advanced classes right then I can start to talk about okay now I have all these containers running around I need some way to keep track of them right and so I talk about orchestration and Kubernetes but that's really hard right there's a lot there and so we have these challenges and so what we decided to try to do was actually can we get a collaborative group of people together to focus on sorry go back one more sorry to focus on kind of these things but we have to talk about them in terms of academia so like every other vertical right you know I worked a lot in sort of like the French services space you know so I know a lot about like options and put you know polls and trades and you know all those kinds of things and I also did a lot in pharmaceutical what I've found since I've been in academics is that they have their own lingo right they have their own stuff that you needed to be able to talk about it in terms of words like pedagogy right which I know had never come up for me until I started getting involved in the university or things like learning outcomes and so we want to make sure that the content is kind of consumable by the people who we want to teach these things so we started a working group basically towards the end of last summer called the higher ed working group it's within the CNCF and you know if you want to check out our charter that's where that QR code goes the QR code will also be at the end so you don't have to grab it now but we have meetings every two weeks and we've started to get some faculty members to show up on the regular and start to actually contribute to some of this content but that kind of leads into some of the challenges which is and I'm going to talk about these a little out of order because the one leads the other so ultimately what we want is adoption we want our content our work to be adopted like every other open source project essentially so that is going to lead us to how we create the thing and then lastly we have also a bunch of challenges around distribution that are not typical for kind of an open source software project most of the ones I've been involved in you know you put down an RPM or you put out a container this has got some other problems so we'll talk about those in a sec alright so first and foremost we have to consider how like a faculty member and when I say a faculty member I kind of mean at a community college or an R1 University like Boston University or at a liberal arts college or whatever in a lot of ways academics are very similar irrelevant of where they're teaching and typically they has anybody ever heard of not invented here that NIH belief there's a lot of that in academics where as as an instructor the default is I don't really use anybody else's things I always kind of create my own which I find really odd right especially now that I've been kind of heavily embedded in open source for a long time I am happy to take other people's stuff and adapt it to what I need right and so but that's not pretty typical so if I shipped a whole class that said this will teach your students cloud native development and you know magic will happen afterwards there's very few people who would adopt it so instead we're going to focus it more on kind of the individual units of a lecture so that we can give people like a piece that they can tie into an existing class or use pieces of it to tie into a new class that they're creating but they're still in a lot of control over the content and then related to that a lot of the faculty members that we're targeting also don't really know about this cloud stuff so in general we kind of have to be sneaky right like we have to kind of do a lot of this stuff kind of on the down low so that we can distribute it but in such a way that other people will consume it like that the people we want to use it will actually be able to adopt it so moving on to the next slide so we have these kind of lecture units I am awful at naming things anybody who's familiar with my work in fedora is well aware of my major project and it's terrible naming so we haven't come up with a good name yet but what we've tried to do is come up with a template for the content and this is and I know you can't read it but it includes all those kind of academic important things okay so something like learning outcomes learning outcomes are often with a syllabus or even with an individual lecture where this is what I as the instructor am trying to ensure that the students come out of it with right and you often won't share the learning outcomes with the students but when I'm talking to another academic it's really really important that we talk about it in terms of what is the outcome we want for the student and when we get there it's a very sophisticated kind of speaking model where we actually want the students or whatever to walk away with a particular set of knowledge and so that's how we kind of rate a piece of like a lecture and so I might be talking to somebody else and say hey these are the learning outcomes I expect for this thing so we include that we also want to include things like in class assignments right or homework assignments because a 90 minute lecture is boring for everyone so it's often you try to put something in the middle to kind of break it up it's why conference talks tend to be 30 minutes instead of an hour and a half long because nobody likes to sit still that long and so you want to break up your class and so we want to make sure that the content is there for them to break up the class so it's easy to adopt as possible and then kind of also the requirements and this is where we get the piece about how do we teach the teacher so we can kind of put in a lot in the requirements of this is what the student should know before you give this lecture and that also informs the instructor of what they need to know before they give this lecture right so we're doing some of those things and then what this also allows us to do is adopt kind of prior art there's a lot of people who've given a lot of talks at conferences in classes in whatever that might be great fits the kind of this open source initiative anybody who wants to contribute we really want your stuff and if you don't have it in the academic ready format that's where we have a few faculty members or even a lot of faculty members who we can kind of add that in so we can go collect that prior art and then kind of encapsulate it in a way that is consumable and then there's of course marketing so then we have to talk about it right so hence for example speaking at KubeCon but in other ways as well we need to go to academic conferences and talk about the stuff there so that we can actually meet our target market alright and then the last thing is distribution so distribution there's kind of all the typical problems of like how do you make somebody aware that it exists but we have weirder problems right for example like a homework assignment I don't want to put all of the answers in public right because I want the students to have to figure out the homework assignment so there's prior art on this there's for example a class out of Berkeley the way they do it is they kind of certify you as a faculty member and then only release that content to you so it's still kind of open source but it's not quite the same thing right it's not just open it's not just available there's other tricks we can do here too where we get more sophisticated where the assignment is actually generated based on the actual student but those are obviously like I said more sophisticated and harder to build but it's this huge challenge we have where we can't just release everything because you know we don't want to give out the answers we want to just give them the questions so that's a problem and kind of difficult in the spirit of traditional open source the other big problem we have is that the almost all the tools involved from presentation tools to you know classroom management tools like Blackboard or Canvas or Gradescope those are all closed source as well right they're all proprietary so it's difficult to kind of release content in a way they can consume because unfortunately to the best of my knowledge LibreOffice and Press has not pervaded everyone right so I can't ship ODP files and expect anyone to use them so this is another challenge is like we need to ship the content in somewhat proprietary formats or figure out ways that we can get around that issue right you know one example might be Reveal but Reveal has its own problems for you know a not technically sophisticated faculty member right so we have a bunch of problems that are I would say atypical for an open source project that make our distribution really difficult but that's kind of where we're at what we've been doing a little bit and what we have next is the kind of you know building more of these units obviously but then at some point we do want to have kind of you know a list with maybe a syllabus that actually says okay here is an actual class and here are all the lectures because it's nice to be able to have a backbone to a class so you understand how the pieces all fit together even if we don't really expect anyone to adopt it as a unit it'll at least give some nice backbone to the content so that we know what like kind of the whole thing and how it all comes together actually go back one and then another thing that we really want to incorporate is experiential learning right so and we should talk a lot about you know kind of that practical tactical aspect particularly about doing a project for a third party you know it's very difficult to come up with projects that you use like in a class that is you know manufactured that has all the same problems that it does the second you have a customer right or a client you know about negotiating scope and things like that which is one of the things that we try to teach in some of those spark classes right so the spark classes they do software engineering projects for third parties it's actually really funny almost every semester we'll have students come to us and be like okay what's the rubric for the project how do I know if I'm going to get a good grade and I tell them I don't know you need to negotiate that with the client and you need to figure out how to make the client happy and when the client is happy you will get a good grade if the client is unhappy they will not this is a very weird thing especially for students at a you know a relatively you know high school where you know everyone's been doing all their coursework all their lives to get you know straight a's or whatever and when I tell them no I don't have an answer for you it really freaks them out but I think it's a lot more like the experience all of you have had in the industry right so that's been really interesting and then kind of the last thing in parallel just the past couple weeks I've been working with somebody in the open SSF who's also involved in the higher ed working group to look at something called like accreditation so it is common for there to be organizations that kind of at a university's program for whatever right and so what they're looking at basically with the Linux foundation and with CNCF and open SSF is how can we say okay and they open SSF if you're not familiar is focused on security so they want to say if you have classes that meet these criteria we're going to give you an accreditation that says you are this student who took those classes is prepared for working in cyber security right what we want to do is actually something very similar and we're doing it kind of collaboratively you take these series of classes and you're prepared to work in cloud native development so it's kind of just another ratification and while kind of on the face of it the student may or may not ever use the fact that it was accredited like to get them a job or something like that what a school can do though is use it to differentiate themselves from other schools to say hey we have a set of classes that will prepare our students for jobs in cyber security or jobs in cloud native development and so that's why we've been pursuing this accreditation activity and really looking forward to that in case you're wondering about these photos this is the if you have seen it the it's often referred to as the Jenga building in Boston this is when it was getting built and I got a hard hat tour early on but so under construction except it's an actual building that I have an office in and so that's about all there is for our talk we had you know we would like to take any questions that you might have oh sorry although it's more like 25 but yeah so the question alright so I'm going to paraphrase for the mic so the question is basically like how did I get from working in industry to work in academia right and so it's actually got a relatively long history mostly related to medicine and lawyers if you think about it lawyers and doctors don't actually have PhDs so technically speaking they actually can't teach at a university so if you want a medical doctor to be able to teach medicine at a medical school they have to have a separate role and that role is called a clinical professor and you've started to see clinical professorships spread out basically beyond law and medicine so for example there's journalism schools now that actually hire journalists as professors but they are clinical professors that's a common one social work is another really common one and you're starting to see it in tech and so my credibility as being a professor is because I'm a clinical professor and so my 25 years of experience is why I have the credibility to teach but I can't be tenure track so oh sure yeah can you hear yeah so thank you very much for your wonderful presentation also congratulations on Lesha it was really inspiring my it will be rather a comment but I think you will add something on top of it let's face one detail you are coming from top notch of academia like Boston itself is collection of most prestigious universities institutions in the world the first city that comes to my mind and I was watching like following your presentation I was thoughtful how can we how can we relate to this actually to academia in developing countries right like how can we bring it up to that level that they can come somewhere near because in developing countries there is actually kind of a dilemma that I know a lot of young people they want to be computer scientists they want to be developers but they believe they believe universities waste of time that I would rather work for four years get four years of experience and I would get better salary better career and so on so how do we challenge that in the end it's a question and thank you so yeah so this goes a lot to the philosophy of it and I would argue that there's a lot of places even in the US that are not Boston right and so there's this challenges dichotomy between like what is the role of higher ed right and a lot of people who are going to you know higher ed whatever you know community college you know liberal arts school university in you know in the US or in other parts of the world it doesn't really make any difference there's this kind of expectation that they're going to the university for to go get a job afterwards in industry say right that's not entirely the goal of a university right what they're actually typically selling is an understanding of the of the science right or of the computer science you know or the political science or the whatever you know it's pretty universal where the reason you're going to that four-year degree is to give you context and then you're doing a master's degree typically a terminal master's degree where that is going to give you the tactical experience that will go and help you into a job you kind of have the other end of the spectrum which is like a coding camp which is basically trying to prepare you purely tactically with very little context and I think both routes are actually kind of tough especially with software development because I don't know that you need five years to be prepped to work in industry so I think that's where we have this dichotomy you know if you were in medicine for example we have several thousand years of stuff that you need to learn over the course of you know typically you know eightish years of degree and it makes a lot of sense but our field it's like a hundred years old right we just don't have that much content and so I think that's where a lot of the disconnect is is that you know you're going for like kind of this practical application but you probably may not or you may not need to do that in five years and then kind of on the flip side of it I think the challenge that we're really trying to address which is not the core fundamental are the people who are buying an education and the people selling an education trying to do the right like are they buying and selling the right thing but more that I actually believe that even in a theoretical scenario you have to have these wider concepts around you know and you see them in some schools right like distributed computing you know my big thing is actually a venture of architectures that you know I think everything should be event driven you know etc that are still not really taught even the theory of it much less the tactical component so I think the tactical component actually gives more context to theoretical stuff as well so that's what we're trying to solve for but I completely hear your point it's just it's it's kind of like it is a problem I'm not sure if it's one that will ever be solved as much as what we're going to see is an evolution of how software developers are created right is that is there going to be different ways to do it and because like I said I think we don't we don't really have a good answer on either side of that house yet you know we don't have enough theory in coding camps and we don't have enough tactical in universities so that is a big open question and not one we entirely want to address sorry back there first of all hard to see you just FYI first of all thank you for coming it's very uncommon to see people from academia to come to the industry events like KubeCon I want to tackle the topic of stubbornness in academia so basically I'm finishing my masters right now at the Delphi University of Technology in the Netherlands and in my department I have two full professors which represent completely opposite spectrums of being open to changes in the industry one of them gave us gave us course where part of it was contributing to the open source code basis for example by making the pull request analyzing the software and then presenting it and another one is actively fighting student organizations who want to work with entities outside of academia like for example big corporations because there might be some involvement of big corporations in academia so it's kind of a purist approach from your experience how apparent it is and how often it happens that this stubbornness could be a problem in modifying the academia curiculum because effectively what we're talking about on the stock is that there would have to be introduced course of release engineering in addition to the software engineering because what most of the universities even research ones are tackling is how to create software but there is close to zero discussion about how to actually make the software run no matter if it's the infrastructure part and the ways of delivering it so from your experience and talking to people in Boston University for example how many of them do you think would be involved in the engineering and how many of them are just too much locked in their academia basement so I don't really want to sing a lot of school but I completely agree with you I've run into both a ton another big topic area that I don't think it's taught very well is debugging is when do you first learn how to do debugging and that's really difficult even though you do it in your class and then try to debug across that how do you even start I think there's a lot of that resistance to change because first of all with good reason a lot of academics don't trust industry and so sometimes they're right and sometimes they're wrong but I think that's a big chunk of it and then the other thing is it's also very difficult to keep up why are all of you here because if you work in software you are constantly learning otherwise you are done you just have to be building building building you can go and choose certain software development jobs where you don't need to worry about that too much but that's I don't know if they're few and far between but that's a very specific choice you're making and I think you see the same kind of problem across faculty is you have some people with the bleeding edge always want to be working with what's new and building new classes and all these other things and you have other people who are perfectly comfortable they're doing a 9-5 job and they are delivering on their 9-5 job but they don't really want to turn it into a 9-6 job right so I don't think that's limited in any way to academia but it is 100% there just like it is everywhere else I also have a question that actually builds up to the previous one what's your point of view of working together with companies I as a company have also tried to collaborate with universities and they're often I get the answer but there are so many companies who want to provide free training so they can recruit the students my point of view is I want to provide knowledge and not recruit students though recruiting students would be nice the primary focus is providing knowledge because all of the companies are complaining about the fact that when someone graduates they don't know anything about what they need to know to actually be able to do the job I appreciate your catch on the correction there so yes right so the challenge a lot of times is a some company wants to provide something that's very specific about the knowledge that they want out of students coming into them so that is often a big challenge and so it tends to be not very popular unless they find a faculty member who happens to really like working in that space or whatever so but I think there are opportunities a lot to provide quote-unquote knowledge but they're usually less formal so like for example our Spark program we take mentors from industry we actually have a lot of instructors from industry but they teach the classes I want them to teach they don't teach a you know like you know I come from red hat right they don't teach a rel class right they teach a class about software development and doing it for clients and if there is some Azure that gets into the mix or GCP or red hat or whatever it's a byproduct of what's needed for the class so I think a lot of people coming from industry in those formal programs tend to be kind of very you know whatever you know focused on just providing whatever their company wants to support but if you're kind of doing it from like I said slightly less informal way and it often still comes up and that tactical knowledge is normally transferable right I mean you know just because I happen to learn some rel stuff in a class doesn't mean I can't apply it mostly except for the awfulness of apt to Buntu and Debian right so I think a lot of it kind of comes through so I think there's there are ways to do that the other big thing that we do and experience a learning particularly in the tech world is becoming more and more popular and but what we we find is that our students gravitate towards working with nonprofits and working with like city government or state government where they kind of feel like they're having an impact on people and so I think if you had a company or whatever who was kind of partnering with like a local nonprofit and doing projects for that local nonprofit but the industry people or the company people right are providing mentorship and things like that I think that can be a better structured way I just think the kind of typical formal mechanism that you see in a lot of companies now you know doesn't doesn't really engage well with kind of the educational concept as much as it's doing you know kind of as you said like work to train them on some tech that said having it as an offering where you know every Thursday you're gonna buy pizza and you know do one of those tactical trainings but it's not a class that's a whole other story right and we do things like that with like tech talks you know pretty regularly where we have somebody from industry come in and give a talk etc and we provide pizza because the only way you get students or engineers at something is by providing food as you all well know so that that would be my kind of thoughts we're done. Oh sorry I need to just add something So I have some experience in academia I want to say that there are professors who are genuinely interested in industry my professor works a lot with Meta so we use a lot of the Meta products and we work on ROGSDB which they've made open source it's not that they keep on contributing to it we also sometimes contribute to it but it works well because we sort of like contribute to their repo and they help us in return so it makes a difference like if you give us something which is attractive to us to just play with like you are developing some software you just give it to us and we just like try it out and if we really like it we actually use it in our research and that promotes your product and that also helps us because we're getting like free stuff so a lot of the things we do you I think a lot of people give us like hardware to test out and then we just try it out in our lab and then we give you feedback on it so it's not always that we are very resistant towards industry tracks because I know like MIT has Tim Kraska who works in databases who's back in Amazon my professor works with Meta there's probably others who work for different companies so I'm pretty sure they're pretty open if you can get the right interest and the right person to get that interest from alright thank you very much and apologies I don't co-present a lot otherwise I should have given more than we should but thank you for coming well over time and I don't want to get yelled at so thanks