 So, good morning. We are excited to share a new program we began at Columbia University this summer to support foundational research computing skills. This program was a true collaboration with campus partners and it is my pleasure to introduce some of those partners who made it possible. Mark Newton is the Director of Digital Scholarship in the Libraries in the Research and Learning Division. Helene Hescock is the Senior Director of Research Services in Columbia University IT. I am Barbara Rachenbach, Associate University Librarian for Research and Learning. Together we will describe this new cross-campus service program to train graduate students in the foundational skills necessary for research computing. We hope to speak briefly today as I mentioned so we have time for questions. I will walk through how we arrived at this multi-stakeholder program. Mark will describe the program and how we've operationalized it and Helene will talk about past and present collaborations between the libraries and IT that made it possible for the work we're doing together today. So for at least four years, the library has been in conversation with campus partners about the need for computational literacy. My guess is on your campuses you're having similar conversations. Too often, we've heard that domain experts, our faculty and the disciplines, were spending valuable course time on teaching research computing skills like R and Python rather than teaching their own disciplinary expertise. To address this issue, the libraries have hired dedicated staff in digital research methods grounded in disciplinary expertise. We've also dedicated spaces, our digital centers for humanities, social science and the sciences where we have hardware and software to support research computing. We've also in collaboration with our partners hosted a number of workshops on our Python onboarding for high performance computing and more intermediate based disciplinary workshops to help in this area. Our campus partners in the Office of Research, the Data Science Institute, Columbia IT have done many of the same activities in parallel. Given these redundancies across campus units, we have long known we needed to partner on a centralized program to address these needs. And we needed to think about how to resource such a program. In 2016, the university ran a boot camp for graduate students working with software carpentry. All told, the boot camp cost the university about $22,000 and we only were able to train about 132 students with a wait list of 78 and no plan to address that wait list. So we knew we needed a scalable solution. This is a question that the research computing executive committee grappled with in their 2017 meeting. To address this question, the research computing executive committee formed a subcommittee on education and training to scope out their actual resource needs for a campus wide program. This is a slide just to give you a sense of the governance that made this program possible. So at the top I've mentioned the research computing executive committee. These are the heads of the major stakeholders on campus, the head of the library, the head of campus IT of the Data Science Institute, the office of research, the engineering school and arts and sciences sit on this executive computing committee. And reporting into that group is what we call the shared research computing policy advisory committee or source pack. We utilize this existing structure to work on a program plan for what we call foundations for research computing. Beneath that you see our two operational committees for our high performance computing clusters and on the right the subcommittees of these entities including the education subcommittee. We also identified faculty who would be key champions for the work in this planning process. So we identified Mark Spiegelman from the earth and environmental sciences to chair this education and training subcommittee. And Chris Marionetti who is in applied math and physics who is the source pack chair. Together they led the effort to create a program plan for foundations for research computing. They themselves have struggled with training their own graduate students on these foundational skills. Additionally, I have to mention our strong administrative partners across campus from the campus IT and from the office of research. So I will quickly illustrate the needs on campus to begin to program plan. We needed to do a gap analysis. So the research, the subcommittee on education and training ran two surveys over the course of last spring. And one was for faculty and one was for graduate students. We have a lot of information about this survey I'm happy to share. But for now, I'm just gonna show you a few key things. We asked the students and the faculty about computational science skills, data science skills, and software development skills. And there was an obvious disconnect between what faculty wanted their students to be proficient in and what students felt proficient in. So in this instance, when we asked about problem solving technologies, you will see the gap that is obvious. The yellow indicates where there is a strong response. So you can see on the left, the faculty felt that the most important problem solving technology from the faculty side was Python. And the students on the right show that they feel that their novices and are proficient in Excel. And then there's a kind of dark murky color around Python, meaning the students just aren't comfortable with that area. So again, illustrating that gap. You'll also see that we asked the students what training formats they thought would be best for them. And it was very clear when you see that bright yellow, it's the pre-semester boot camp. So the kind of concentrated boot camp approach. So therefore, aimed with data from the surveys, a program planned in a budget. The Education Subcommittee made the case for Foundations for Research Computing at this last spring's Research Executive Computing meeting. And it was approved and funded for a year at that time. You'll see our funders listed on the slide. We have a budget of about $200,000, half of which comes from a dedicated resource that the library has put forward as their contribution. And I am happy to say we just hired a coordinator who will be starting in early January. And this is something the libraries felt strongly that we wanted to have a coordinator in the libraries to acknowledge that this is a new form of literacy, computational literacy. And we felt we were the appropriate place on campus to coordinate. The other 100,000 is to enable us to be members of software carpentry. So we have an institutional membership. And we are also running a number of other programs associated with the carpentries that Mark Newton will tell us about in a moment. Thank you, Barbara. And so now I'd like to tell you a little bit about the next chapter of this work, how the program, Foundations for Research Computing, this collaboration, this pilot that we funded as an experiment for one year has preceded. And where it appears to be headed now that we're about halfway through our planned activities for this year. Foundation seeks to cohere, brand and elevate new informal instruction that fills those gaps that were identified in some of those earlier surveys. Filling the niche in between the do-it-yourself opportunities that's offered to students through the licensing of the library's electronic resources. And then on the other side, the formalized for credit courses that are extended through the schools. The foundations program also presents us with the perfect opportunity to embrace the significant alignment with our faculty and with our campus administration who've come together to identify this instructional need and we have some pretty clearly defined supports to help get us there. First, by identifying the high level aspirational targets that have served as our touchstones, that we want to enhance the instructional supports available to step students through instruction and get them up and running with the basic skills of data science. Everything from working confidently with their personal computers at the command line to getting up and running with the plotting and programming and statistical libraries that they can take into the classroom, into the lab, into their own research. And that we can focus on the alignment question through this system of light governance for the foundations program itself. Our key faculty champions and senior level administrative stakeholders occupy, they comprise an advisory committee that monitors the progress of our initiatives, helps us to identify opportunities for expansion, and define program requirements that then get executed by members of a coordinating committee that's charged with the oversight of the operational and the logistical details of the program. In some respects, I think this model of oversight is a version of the very robust picture of oversight that we saw in place during yesterday's collaboration presentation, quite impressive from our colleagues at Yale. And at the end of the academic year then, it's that the chair of the advisory committee will report on the program outcomes to the research computing executive committee that charged it, and to which we hope will be successful in securing funding for subsequent cycles of this program. The plan that came out of the RCEC had already established and specified specific elements of the program framework, its essential components. The centerpiece, of course, is these pre-semester two day boot camps that are built on the established and trusted pedagogy of the carbon trees. I think many in here are likely to be familiar with these open source curricula that are published, maintained, and expanded by that membership organization. The carbon trees provide a means of guiding novice users through a tested approach to the introduction of the key topics, and then perhaps even more importantly, of helping our novice instructors get comfortable and familiar with effective ways of delivering materials in the classroom. Beyond these instructional courses or boot camps, there's the workshops, there's peer consultation, we have distinguished lectures from visiting speakers, and we're planning for a research symposium. Doing everything but stopping well short of what you see in that advanced category, there would be overlap with courses that show up in the schools themselves. So for the libraries, which as Barbara mentioned, were ecstatic to offer an institutional home and a staff line to help coordinate this program, there's still the question of where within the libraries to situate those efforts. The emergence of foundations came about at just the one year mark after charging a new unit focused on the extension of digital scholarship activities supporting research and learning across campus. Among our digital scholarship or DS team, I think you'll find a number of pretty recognizable activities, open access repository support, scholarly communication outreach, scholarly publishing services, digital humanities community building, digital project development. Even we have a technology lending program, 3D printing services. And to this, we now add the coordination of the foundation's program, which helps us to further fulfill our mission in both research and learning through direct support for the development of course, skills that catalyze discovery and inspire inquiry. This is language borrowed from our libraries on strategic directions. And we knew we wanted to bring the coordinator into the libraries, ASAP, but the program schedule itself lit a fire that had our multi office project team working just right from the get-go. As Barbara mentioned, the foundation's program was green lit this past summer for rollout in the fall. Nobody panicked, but we did some pretty what I'm calling orderly scrambling, facilitated by the collaboration aspect and the essential but distributed business knowledge that's held in personnel across the libraries and research initiatives and IT. That way we could brainstorm logistics, clear some administrative hurdles and extend the boundaries of the trust that we have with our faculty. And it's hard to see how this could have happened otherwise. We were charged, resourced in June. By July, we had moved allocations into a working budget. We'd begun processing the carpentries membership through our campus IT. We'd identified six brave volunteers among the full-time staff members of the libraries and IT full-time staff to become certified carpentries instructors. We'd launched a descriptive website. We'd issued email to 20,000 graduate students among the among the disciplines and the professional schools across multiple campuses to participate in these two consecutive full-day bootcamps. And it turns out they were really, really popular. Within literally four minutes of the registration period opening, we had filled all 90 of our initial seats for these foundations for research computing bootcamps. The wait lists grew to several hundred long immediately thereafter, which I think told us two very important things. One, clearly that we had tapped a nerve, that there was immense appetite for this kind of instruction. Yes, the program is free to participants. And yes, we fed them, but the enrollment served as a real commitment. Not only were participants, not only were our classes filled, but the students stayed overwhelmingly engaged through dense material, most of which was entirely new to them over the course of two full nine to five days. Our pre and post surveys gave us feedback about the knowledge of our instructors and on the utility of the programming. Two, it told us that the demand was much more widespread than we had originally envisioned. Foundations for research computing is targeted toward graduate students in research programs who we imagine have immediate or near term opportunities to apply what they've learned. But in the demographics of our attendees and our wait lists, we found widespread demand expressed in the professional schools and continuing education and beyond. Our programs email alias received numerous inquiries, requesting expansion of the program for postdocs, for research support staff and other members of our campus. Enterprising students even waited outside our welcome tables on the day of the boot camps, hoping to fill a seat from a late dropped student, and I have to say at least one or two of them was successful. Across all the other program components thus far, it's been a little bit more of a mixed bag. The distinguished lectures bring influencers and thought leaders to campus to engage students, but they also have the knock on effect of providing these periodic regular promotional promotional boosts to the program itself and all of its related activities. And I think they've been pretty well received. They're also a very visible product of the collaboration of the partners as IT library and research office staff, some of whom you see on this very stage are the ones who are staffing and vigorously promoting these events. I think the workshops too are highly emblematic of our collaborative efforts and they now benefit from a unified marketing and communication strategies supporting the workshop series and other classes. For example, on HPC out of my colleague Helene Hescock's groups on statistical analysis and research data services and more that are offered by the different library divisions. Hopefully all of this to greater effect. The office hours concept in which students with extensive subject knowledge in data science, programming and other related topics staff a drop in hours for peer consultation at distributed locations in the library system each week has been somewhat less impactful so far. We've conducted marketing and awareness campaigns for these services but the uptake has still been comparatively slow to catch on. And this is one of those areas where having governance structures in place for foundations is proving especially useful as we're as we're working with them to help us think about evaluating and adjusting our efforts midstream. We have just enough experience under our belts at this point to begin to to piece together the elements of on program effectiveness that we're hoping to report back to the research computing executive committee at the end of the academic year. We are extremely fortunate to have available to us access to the library's in-house assessment team who are helping us with surveying and data collection that will feed into such reporting at many points along the way and everything that we're doing at this point is being done with a view to helping to build that narrative back to the committee to help with the sustainability of the program. The excellent news as Barbara mentioned is that we have just hired our full time program coordinator who will be helping us and be a very welcome set of helping hands going forward. I think we can all say from direct experience how ecstatic we are to have this person aboard and how essential this role is going to be in supporting the program that is clearly has such growth potential. With respect to activities post the August boot camps we've trained and certified another six instructors. We did this in August. We're planning to host another 120 students in boot camp sessions this coming January. We're expanding the roster of distinguished lectures from three to four for the spring and we're going to open a regular sequence of much smaller workshops through the spring to augment the boot camps and to help keep students engaged and their skills refreshed along the way. We're also beginning to reckon with some of the emergent questions that have surfaced as we've dived headlong into this project. Both our medical campus and the school for professional studies present as particular opportunities for us for expansion as that demonstrated as demonstrated by the interest in the enrollments that we saw this past August. And this has opened doors to discussions with administrative leadership in these schools around the potential for growing the body of stakeholders in subsequent years for the program itself and for foundations to support the training of additional instructors in that would then deliver school and discipline specific instruction back to their home programs thereby scaling foundations in a way that would be difficult or impossible beyond the bandwidth of our initial instructor cohort. As I mentioned, we're taking a hard look at the usefulness of the drop in office hours sessions led by the peer consultants were likely to pivot again to these more active workshops that focus on discrete topics that can be promoted individually across the campus events calendar and social media, etc. And finally, like I say, we're thinking about how best to direct and interpret the data that our assessment team is helping us to collect. It's important to the vitality of our program to have data that help us help our faculty tell the story quantitatively that I think all of us are feeling anecdotally as we build out the field of participants and grow the community of student learners, faculty proponents, staff instructors and peer consultants that embody the foundations for research computing community. But it's even more important to know for ourselves that this work is having the intended impact that that establishing the through lines that show that participation in our program is a net positive for Columbia graduate students still remains a fairly tricky proposition but one that we know is absolutely essential going forward. So some questions here that don't have absolutely clear answers, but that's what makes the wealth of experience that's available to us through this collaboration, the library's IT Office of Research invaluable. And so with that, I will pass now to my colleague, Haleen Hescock, the senior director for research services at CUIT to talk more about the development of these kinds of collaborations from her perspective. Thank you. Thank you, Mark. Okay, so I'm going to just take a little bit of a turn here. While this is focused on the foundations program, what I'm going to be talking about is really the collaboration between the libraries and IT. I'm setting this on top of the computer here. So the sorry. Yeah, you can take this. Okay, so that's that's good. Thanks. So we're going to talk about the collaboration between us and it's not just the foundations. The foundation is the most visible and most, you know, the biggest collaboration that we have right now. And but it's not the first one between the libraries and IT. And not all Barbara talked about the governance structure of the research computing policy advisory committee and not all of these programs that we've collaborated on fall under that, but we do leverage the existence of this committee for feedback. We go back to faculty. What do you want? What do you need? How can we serve you? So as they said, my department is research services and I'm in central IT. I'm not in libraries IT. And I support research computing services and you can go to the next one. We support research computing services and research administration research compliance applications as well as research computing. So everything that my group does is is to serve research, not the libraries. But this is where our worlds collide or intersect. So this is how we how we get together. One of the services that I provide is a shared HPC service. We build the clusters. We provide technical support, system administration, as well as providing some HPC training each semester. So it was natural for some of my staff to become trained in the carpentries and become part of the foundations. It was also natural to roll our HPC training into the foundation. So this all becomes cohesive. In particular, the carpentries we ask for volunteers in central IT to be trained as trainers. So it's not just my staff, but we have we have buy in from our CIO. They realize what an important program this is and they support anybody becoming part of this program. They need managers approval and because it's a big time commitment to be a trainer in the carpentries. And we do ask that anybody we train and get certified commit to at least two boot camps a year. So these are some of the first slide was just some of my services, but these are the collaborations we've been doing with the libraries. And these have gone over the past few years. So we we get together. What are what? How do we complement each other? What are the synergies? What can we provide? What services can we provide that the researchers need? The libraries can help guide the students to the services. And I would say we do the tech and they do the they do the talk. Our first collaboration was the implementation of electronic lab notebooks. And that was started in 2016. And to date, we have over 2,600 people using this platform. And we we split the cost of the licensing libraries and I.T. and we split the administrative support. We work on strategizing and marketing and communications and direction. And we all meet regularly and work together on that. And we also involve the Office of Research and we meet with them. They're also part of it, even though we're paying for all this. We list our shared resources on our respective websites. We do we have our own mailing lists and our own target audiences. So this helps to broaden the reach as we all work together. We're also working on a pilot right now for a computational reproducibility platform for researchers to share, publish, collaborate code. This is a service by Code Ocean and it was brought to us by the Office of Research as a platform that they thought there was a real strong need for. So we took that on. Again, we're splitting the licensing cost. And you notice how often I use the word we. Because we are doing this all together. I just wish they were closer than 15 blocks away because I'm running over their offices all the time. And I think one of the reasons we have the successful partnership is that we all feel like we have a stake. So these these shared these shared services are not listed as a library service or an IT service, as I think this was mentioned in the keynote as well. But they're listed as research services and role working together on that success. And then that's about it as a side note as Barbara and Mark invited me to come here. This is I didn't even know what CNI was. And this is a really great group and I probably will be coming back. Thank you, Barbara.