jonathan@researchcomputingteams.org
In this interview, Matthew Smith from the University of British Columbia and I talk about managing as a new manager as the pandemic hit, finding other teams to work with when we can’t do it all, the benefits of communities of practice, the self-doubt that can arise managing more experienced employees, and a coaching approach to being a manager.
Some key takeaways:
On coaching:
Diving into the coaching based approach to leadership has been of value in the midst of the pandemic. […] Because yeah, it’d be a rare bet where we’re gonna have circumstances where I’m the technical expert, amongst the rest of my team who collectively have over a hundred years experience. Coaching is a skill, and that’s where I think I’ve been able to, to bring value to the team is providing space to bring their problems, to focus on a problem or solution they wanna reach and then offer a chance for them to see my perspective or other perspectives on how they might approach a problem, whether it’s technical or interpersonal.
On a very rapid increase in workload at the start of the pandemic, and helping the team by saying “no”:
Our biggest challenge [with the onset of the pandemic] was we saw our usage double or triple overnight. We had researchers hadn’t even heard of ARC, but they wanted to continue doing their research. And a good portion of that happened within few months of the pandemic becoming a reality. It was easy because the team is really passionate about it and, you know, it was a bunch of new friends, new researchers, new people to learn about and understand what they’re trying to accomplish and help them find the solution they needed. But at the same time, we were kind of not becoming aware of how long we were sitting at our desks and we started to feel a sense of burnout after not too long. So being able to step back, recognize that and figure out how we could maybe work smarter about delivering all our services was a challenge. […] [I had] a willingness to be their bad cop and make that known. I think knowing that there was someone there to kind of offer a "no" in a way that wasn’t gonna put the onus on them helped.
On not trying to do it all by enlisting the support of other teams:
It seems there’s never enough resources to be able to provide everything that researchers need. Our researchers that top in the world and that comes from being incredibly ambitious. They’re not limited by their ideas, but by all the different components to make their ideas for their research study a reality. So [..] it was all about trying to stretch the dollar. To find different avenues or partners, etc, to work with to ensure that we’re focusing our efforts on where it makes sense.
I really enjoyed the conversation, and I hope you do too!
Jonathan Dursi: Hi! I’m here with Matthew Smith from UBC. Matthew, why don’t you tell me about yourself?
Matthew Smith: Sure. Thanks for having me here today, Jonathan. So, yeah, I’m currently the manager of research systems for advanced Research Computing at the University of British Columbia. And for those who aren’t familiar with what we do here, we are the central research computing unit for both UBC Vancouver as well as UBC Okanagan.
We look after all central research computing infrastructure for the university including on-prem as well as some peripheral cloud technologies. My team specifically is the research systems team. So we are that, that infrastructure focused team. Our bread and butter right now is our on-prem infrastructure as a service offering, which we provide through our cluster.
We affectionately call UBC ARC Sockeye. All of our on premise infrastructure is named after fish. And that is a 16,000 core traditional HPC architecture. And then we have about 1.3 petabytes of scratch for high read/write capabilities, and then another one and a half terabytes of project.
In addition, we do offer cloud slash object storage where we have about 10 petabytes right now, that we offer to the research computing community and no charge to the, yeah, to UBC researchers.
Jonathan Dursi: Very cool. Are all of your researchers at UBC or are there external collaborators
that work with you too?
Matthew Smith: That’s a great question. All of our services revolve around a, what we call an eligible UBC researcher. So essentially a faculty member or a faculty adjacent position that holds. So we do support the work of, of lots of collaborators, but there’s always a UBC researcher at the helm and kind of directing things.
So to get access to our system. It’s as simple as a UBC researcher signing up, getting access, and then they can provision access to that capability, data, et cetera, to whoever they would like.
Jonathan Dursi: Matthew, how did you get started as a manager of your team?
Matthew Smith: So where I’ll start?
Yeah. I started here in, actually it’ll be December. I’m about to hit my third year anniversary, so it’ll be December in 2019 just before the pandemic. Yeah, I’ve been, you know, a huge fan of the work that ARC does and the work that, that my current team has done. Yeah, there are members of my team that are pioneers in research computing, and, and I’ve been, I’ve been on campus here for almost 14 years, and I’ve been a fan.
I’ve been privy to their work and what they’ve been able to accomplish for the research computing community. So when I saw the position come up after going through cycles of self-doubt, I went ahead and applied for the position. Had some conversations with, with the folks at ARC as well to make sure what I was going to bring to the table was gonna be of value to the team and yeah, the rest is history.
Jonathan Dursi: Very cool. I’m glad you managed to overcome that self-doubt. It’s, our field is so vast that our own personal knowledge of it can only be a fairly thin sliver.
And you were in sort of the broader research computing and data space before this position
too, right?
Matthew Smith: Yeah, that’s correct. So I had a, a couple year stint with the faculty medicine when I first started at UBC back in 2009. But after then I left and went to the UBC Psychology Department.
Which is the largest research department on campus. There’s some other large units within medicine, but psychology is all one department. Without any fragmentation. And it’s also Often the amongst the top five psychology departments in the world, and number one for developmental psychology.
So we can think of things like baby psychology, how the brain develops, et cetera. So I, yeah, I started, my role was to provide desktop support for Linux users as my primary function. However, after being there a year or two, it became clear that there was this huge need in the research computing space.
So, you know, working with my department, head and director of administration, we refocused the IT unit within psychology’s efforts on research computing and look to essentially leverage our partners with both faculty, IT, and central IT to provide kind of the more foundational IT services while we, you know, worked on the, the crazier
Well, yeah, the, just the the more dynamic research side of things that just seem to, to warrant a little bit of site specific expertise and focus.
Jonathan Dursi: I imagine within a department that large, even just one department, you probably saw a fairly diverse range of use cases. Is that right?
Matthew Smith: Yeah, absolutely.
We had at the time. It’s grown since then, and I supported 55 different research labs. We had, I, I wish I could give you the number of areas, but yeah, we had folks ranging from social psychology to child psychology, clinical, et cetera. So yeah, while we could, you know, often find some commonalities to, to launch some central services, it was quite diverse.
And yeah, we did have. Certain services and platforms focuses, et cetera, that were specific to those areas.
Jonathan Dursi: Cool. Cool. And so either there, or when you started at ARC what were some of the challenges you faced when you started managing a team? Or even as a lead, as the psychology team sort of grew around you?
Matthew Smith: Mm. Number one, that’s probably the same across the board as part of resource constraints. It seems there’s never enough resources to be able to provide everything that researchers need, Like, you know, our, our researchers that top in the world and, and that comes from being incredibly ambitious. So they tend to be, you know, not limited by their ideas, but by everything that, you know, all the different components to make their ideas for their research study a reality.
So with psychology, it was all about, yeah, trying to stretch the dollar. You know, find different, different avenues or partners, et cetera, to work with to ensure that we’re focusing our efforts on where it makes sense. And then I think that carries a little bit into ARC while we do get a lot of support for our VP Research to be able to support research in a big way.
There’s still never enough. And that’s been kind of ARC’s mission. We are the central research computing unit and we have teams that provide other services beyond just infrastructure, but focusing purely on infrastructure. We also compliment what’s available through the national system now known as as the Alliance.
Right? Which, which they provide these large, almost. High performance computing systems and storage systems. But you know, prior to ARC starting up, researchers could experience anywhere from, you know, five days to a two week wait time. So, you know, a main. Part of our mission was to provide supplementary resources to reduce those wait times.
And then we’ve continued in that and looking at what gaps exist beyond just resources. Like is there you know, things like training platforms that make it easier for, for folks to leverage high performance computing capabilities and, and other pieces like that.
Jonathan Dursi: Yeah, that’s, that’s terrific. When trying to choose priorities. When you can’t do everything, when you don’t have the resources for everything. Finding those partners and, and not trying to do everything yourself is, is so important. I’ve seen teams burn themselves out trying to do everything when there’s, there’s other teams that can help.
Matthew Smith: Yeah, a hundred percent.
And yeah, the team here, I’m honored to be able to support the team here. I have many pioneers of research computing at, at UBC here in the team. And then just lots of brilliant minds and they’re always finding ways to say yes t o researchers. So finding that balance and, and being able to, to build the solutions that we know we can build while also trying to be mindful of the time that we have available in a day is challenging.
Jonathan Dursi: Cool. So tell me more about the team and the sorts of things they do day to day.
Matthew Smith: Yeah, absolutely. It really varies. So thinking about numbers I support seven members of the research systems team six of which are based here in the UBC campus. And then we have one of our one of our folks that is based in the Okanagan. And my team consists of three architects, so we have.
I’ll, I’ll go ahead and, and name names. And so we have Roman Baranowski, absolute pioneer of research computing here at ubc. He built the the first large scale supercomputer here on campus, and he’s been, he’s been doing what he loves here for, for 25 plus years. So he’s our software architect. And he knows the, Yeah, he, he’s the brain of how the work, the workloads get distributed.
And essentially manages our, our resource scheduler, that, that makes sure everybody has fair access to the system. Then we have Ryan Thompson, he’s our systems architect, and he looks at how everything connects together and ensures that, you know, our planning and our infrastructure meets operational needs.
And he works with the other, other teams at ARC as well to ensure that again, that, that continuity of service to make sure that we. We can deliver on the back end, what folks are, are planning for the front end. And then our newest addition to the architect world well to the, the architect classification, but certainly no stranger To UBC research computing is Venkat Mahadevan. He is our cloud architect.
Jonathan Dursi: Nice.
Matthew Smith: And he’s been basically looking to leverage not just what we think of in cloud in terms of AWS, Google, et cetera. But, but cloud technologies in general. So kind of looking at a little bit more composability than we’re used to with, with traditional research computing.
Then architects aside, we have the rest, rest of our folks are system administrators. So I can touch base on, on those, those experts as well. So we have Wade Klaver that’s been based in the Okanagan and Wade’s. Yeah, Wade’s been part of the research computing mission along with our our director Steve Cundy.
Since the beginning. And he was involved in the formation of arc, which I think first started around 2015, but became a, a formal entity in 2017. And Wade is our, our connection to the research community. So he has happy hours in the Okanagan where he invites researchers along to learn about ARC and just trying to talk casually about how we can support the research. And he’s, yeah, he’s, he keeps the social side of things going along, along with Roman, who’s a social butterfly as well. Then we have Ken Bigelow. Ken’s been working in research computing here at UBC for over 20 years specifically with brain sciences. And again, Ken has all the team possesses this ability.
But, you know, he, he can explain complex technical scenarios to. To, you know, anywhere from non-technical to a highly technical user and is able to kind of gauge how we go about that. Then we have Jacob Boschee. He holds a PhD in physics with an expertise in quantum computing, and he is our bridge to the ever growing world of, of quantum simulation.
Jonathan Dursi: Nice.
Matthew Smith: So not only is he an expert systems administrator but he can work on specific solutions for the research community. Given the development of our Quantum Matter institute and, and their kind of rising presence in research here at UBC, we’re grateful to have ‘em.
And then last, but certainly not least we have Mykhailo Kutsenko who, y eah, again, has over 20 years experience in, in computing specifically a passion for, for delivering services for scientific research. And Misha is our, our man on the ground. He understands the actual functioning of the hardware and spends almost every day in our data center ensuring that, that all the hardware gets the TLC that it needs and reserves.
Jonathan Dursi: I love how even the short description shows how well you know each team member and, and their strengths and interests. That’s, that’s fantastic.
Matthew Smith: Thanks. Yeah. No, I yeah. Every day is, I learned something new about them which yeah, I’ve been working closely with them for three years and we’ve experienced a whole worldwide pandemic together, but we continue to, to, to learn and, and grow together, which is awesome.
Jonathan Dursi: And three years, that’s, that’s barely enough time to establish yourself before the pandemic happens.
Matthew Smith: A hundred percent.
Jonathan Dursi: How, how did you manage that, especially with. From the description of your team, you, you have one person who’s already a little at risk of being isolated by being at a different campus.
How did you handle the, the switch to remote?
Matthew Smith: So, yeah, you know, I think we, it happened pretty easily for us, and I think the. Challenge was really our time and trying not to burn out. Right. You know, with regards to the gentleman we have in the Okanagan, you know, we always had our, we had a hybrid video conferencing room, so any meetings we had, they would be present there in video conference.
But there was this kind of dichotomy of those in the room and those who weren’t. Yeah. Because we were having kind of, you know, some side conversations over, over the table that they’re probably not able to understand. And there’s also that, that time delay. You know, no matter how slick your your VC system is Yeah.
Is, is present in a conversation. So I would almost say that kind of improved with the pandemic. Basically, we were told within two days that we had to switch to to full remote. And what that meant for our, our remote participant is that. He was now kind of equal with the rest of the team in terms of how we met and, and conversed.
So and, and thankfully we’ve been able to retain that. So even though we’ve moved back to, to being on campus two to three days a week we continue to hold our team meetings purely on, on Zoom with, with multiple folks in, in their own, in their own rooms. Just to have that, yeah, maintain that. Our biggest challenge was we saw our usage.
Like double slash triple overnight, where we had researchers that, you know, hadn’t even heard of ARC, but they wanted to continue doing their research. They were being prevented from coming into the office and, you know, performing their lab work and they. You know, we’re aware that their colleagues were working in this, this wonderful virtual space that is research computing.
So we had yeah, between before the pandemic and now with the current levels we’re at, we’ve seen our, you know, the amount of researchers using our system tripled.
Jonathan Dursi: Wow.
Matthew Smith: And a good portion of that happened within Yeah, a few months of, of the pandemic becoming a reality. So being able to, I guess keep up with that, that rise in demand.
It was easy because the team is really passionate about it and, you know, it was a bunch of new friends, new researchers, new people to learn about and understand what they’re trying to accomplish and, and help them find the solution they needed. But at the same time, yeah, we were kind of not becoming aware of, of how long we were sitting at our desks and we started to feel a sense of burnout after, after not too, So being able to kind of step back, recognize that and figure out how we could maybe work smarter about, about delivering all our services was, was, was both a challenge and something that we were able to, to overcome and, and take on.
Jonathan Dursi: And were there particular things that you in your role as a manager did to sort of help that realization come about and help that reprioritization or. Tweaking of, of the effort to make things a bit more sustainable.
Matthew Smith: I think it was you know, a willingness to, to be their bad cop and making that, that known.
But, you know, just kind of getting out there and saying, Hey, I, I know everyone’s busy, but, you know, I’m there to basically have a, a second eye on things. If you feel like you know, too much of your time is, is going to be invested in something and then. Proactively taking a look at some of the tickets that were coming in and, and asking my team how I can help with situations that I, I could tell were going to be a little bit resource intense.
Yeah, they were already, you know, great about leveraging each other’s expertise. But I think knowing that there was someone there to kind of offer a "no" in a way that wasn’t gonna put the onus on them.
Jonathan Dursi: Right.
Matthew Smith: Was probably a little, a little helpful. Basically just knowing that they had support to step away from things when it made sense.
Jonathan Dursi: Fantastic. You’ve talked about leveraging expertise within the team, and you’ve mentioned the broader ARC group. What, what are some of the other. Teams that your team works with?
Matthew Smith: Yeah. So maybe I’ll focus on ARC and then we’ll branch out from there. So we’re the research systems team infrastructure and, and also recognizing that research computing was infrastructures of service for decades, right?
So my, my team also has a lot of historical knowledge of how things have progressed over time. And then we have a research platforms team. So they look after software platforms and basically look to address that ease of use piece. They work closely with our architects to, to develop those solutions.
But most recently that’s been ramping up with our cloud architect because… I think there’s a close tie-in between c loud technologies and a lot of these ease-of- use platforms. You know, serverless is a big buzzword right now and that tends to want back onto Kubernetes, which of course a cloud technology.
So yeah, they, they look at they launch basically web GUI, software, GUIs, et cetera. We currently host REDCap. That’s, that’s their bread and butter right now. But we’re looking at growing that exponentially. Over the next year or two. We have quite a, quite a bit of work in the hopper. Then we have our sensitive research team.
They look after all the privacy and security pieces. And this is where, yeah, my team is, is really grateful to have their support. Because when a researcher has a question that has to do with, you know let’s say policy, compliance, technical controls, et cetera, we don’t have to address those directly. We have a sensitive research team that kind of liase between our team and the researcher or.
Governing Body health authority, et cetera, to basically, you know, handle that on our behalf. Then work with us on solutions as needed. And then lastly, but certainly not least, we have our research specialists and they are our training team. So they look at, they kind of work with all the teams including our, our support analysts.
So we do have a, a help desk within that software facing team. And so they work closely with those folks to understand. The types of issues that are being trouble shot the most, you know, where the pain points are and then develop training solutions to, to address those proactively. Terrific.
Jonathan Dursi: And oh, sorry, were there other teams you were going to describe?
Matthew Smith: I was gonna extend it out to basically central IT and talk a little bit about the relationships there. So, you know, we do manage our on-prem, equip on-prem equipment as well as cloud. But everything connects to. The back end, that is UBC. So folks log to all of our services with their campuswide login, which is a single sign on provided by UBC Central.
It, so we work closely with their identity and access management team to basically ensure that we’re on top of you know, the latest changes when they come out. And that every time we launch a new service, we do need to engage with them to ensure that, you know, our timelines sync up with theirs and we’re not, you know, gonna introduce any unnecessary bottlenecks or, or time things in a, in a manner that doesn’t work for them, because we do need to, to work in tandem with them every time we, we offer a new solution.
And again, that that goes back to. You know what I found success with in working with psychology, I was delighted to kind of see like minds around that when joining. ARC is this idea that if we don’t need to, to invent a new wheel, let’s let’s leverage the work of our colleagues. Same goes for network infrastructure.
So we do have our own, you know, switch back-end, et cetera in the data center. But we leverage UBC IT’s uplinks and work with their network management center, their network engineers, et cetera. Basically, you know, take a look over our proposed architecture, help implement it, help troubleshoot issues when they come up.
And, and yeah, I’m proud to say that we have a, we have a very good relationship with our central IT. We wouldn’t be able to, to do what we do without them.
Jonathan Dursi: That’s great to hear that. I hear from teams who, who have somewhat more fraught relationships with their central IT. This is, this is great. Do you, does ARC and central research computing, does that report up to the CIO or up to VPR, or,
Matthew Smith: Yeah, that’s a great question and we have a bit of a unique relationship there.
So all art staff, including myself, everyone except our director is under the VP research portfolio
Jonathan Dursi: Interesting.
Matthew Smith: Our director reports to both the VP research and the CIO of UBC.
Jonathan Dursi: That’s really interesting
And I think, and your director has, has been at UBC for, for some time to really build this relationship. Is that right?
Matthew Smith: Yeah, over a couple of decades. He, he’s based in the Okanagan and he was part of UBC Okanagan. Before it was UBC Okanagan, it was Okanagan College and then Okanagan College split off into, into the college and the, and the university that became UBC Okanagan. And he was at the helm of ensuring there was research computing capacity for the Okanagan.
That’s yeah. Since day one.
Jonathan Dursi: Wow.
That’s great. And so because there, because there is a central research computing. Are there still individual, say research software developers scattered around campus? Do you, do you work with some of these smaller teams or, or even solo practitioners scattered around?
Matthew Smith: Yeah, we do.
So there’s a bit of a difference between us and UBC Vancouver and the Okanagan. Sure. In the, in the Okanagan, every bit of IT support is centralized. Recognizing that there may, you know, be a, a tech savvy lab member, postdoc, et cetera, that is performing some degree of operations in the Okanagan.
They’re doing it in tandem with research computing here in Vancouver. We provide, you know, central services for advanced research computing. But let’s say you have a desktop and you wanna install some research software. Or you you have research software installed that’s not working as you expected, trying to replicate a study, et cetera.
Here in Vancouver you have to get that support from either your department or your faculty. Most often the department and then some of the research centers on campus, indeed do have their own technical experts that they hired to be part of their, their lab or research center. So yeah, we do, we work with, we work with everyone, and that’s part.
You know, I think something that members of my team have always done is network directly with those, those individuals. So UBC does have a systems and network administrator group that we are part of. It’s kind of an informal gathering of folks, there’s a slack, et cetera. So, so we take part in that community and then leverage, we call it the it UBC backbone.
That’s a, that’s a bit of a funny way of, of referring to it, but essentially it’s all of our. Our friends and partners that, that yeah, provide it like, or research it, like services on campus. And indeed, it’s been those connections that I think have led to our success in just being known on a campus this large.
Jonathan Dursi: Sure.
Matthew Smith: UBC, we have, you know, about 85,000 community members. So being known to all those 85,000 people isn’t something that you’re going to accomplish. You know, throwing a, having a sign out in front of your office? Often we have, you know, members of labs departmental it, et cetera.
Those are the folks that are recommending researchers come to us.
Jonathan Dursi: Cool. Could you tell me more about the, the community and the Slack connecting these different groups? There’s lots of certainly members of the research computing teams community that have tried to bring together sort of communities of practice.
And sometimes it’s successful, but often it’s, it’s a lot of work to try to bring them together. How, just mechanically does this community within UBC work now. Are there things you think that were really successful in bringing it together?
Matthew Smith: So the system and network administrators group, it’s known as it’s acronym here on campus as SNAG. And it’s existed long before I started at UBC. I get the sense it’s been around for a couple of decades and it started as a pretty barebones mailing list that was , I think it was created by the IT folks from Population and Public Health.
So, I believe it was a research IT unit that, that created it. And the idea was, yeah, to, to basically build that community of, of folks that were engaged in, in similar practices, whether you’re in central IT or or a department unit. Also UBC was more fragmented back then as well. A lot of the units that now make up UBC IT were their own units.
So there was like a web team that were their own thing. An AV team that was called, like the media group that was their own company.
Jonathan Dursi: Mm-hmm. .
Matthew Smith: And so we, it started as a way to connect all of those dots and then basically it’s been a mainstay that hasn’t really wavered. The mailing list still exists but most of the conversation moved to Slack.
I think about five years ago if I had to guess. Interesting. Yeah, just that’s the quick answer.
Jonathan Dursi: And the groups I’ve seen that have been successful pulling something like this together uh, You know, you can’t make people be participants in these things, so there has to be something the participants are getting out of it.
What, what would you say the value is of SNAG to the people who participate?
Matthew Smith: I think hearing about things through unofficial channels and that’s, you know, UBC IT use SNAG as a way of advertising, you know, events they’re putting on, or even kind of broad let’s say advertisements, but things that are going on at campus that might impact others operations.
We do have, like for cybersecurity, UBC IT cybersecurity have their own portal, but it’s a way to even draw attention to like a post on. So to, to let the SNAG folks know that, hey, there’s something you should, you should catch up on. And I think it’s always been a place that has fostered honest dialogue so, Folks, because we recognize that, you know, there’s no faculty members in there.
There’s maybe a couple faculty members that also do serve as technologists, but it’s, it’s all technologists there. So folks speak their mind and you know, it’s not the type of forum where you’re, you have to worry about any sort of recourse from saying the wrong thing on there. I think you know, UBC IT leadership know that that’s SNAG.
And it’s a, it’s a place that, Yeah. That folks kind of talk on honestly, Right? And and engage directly. Yeah, I think in the past, pre-pandemic SNAG would also organize like presentations, et cetera. Although we haven’t seen as much, I think we’re just starting to kind of get back to, to some of the more in person things, purely because the, the online stuff works so well in, in terms of, of a distributed environment like we have at UBC.
In addition to SNAG, I’ll point out that there is what’s called the General Research Computing community practice that operates in UBC. It’s a smaller group of folks that are. More focused in research computing at that kind of middle scale. So they’re not operating large supercomputers, but they might manage, let’s say 10 to 12, 10 to 15 node clusters.
And so we do have we have members of ARC that, that attend those community of practice, but it’s led by members of the different research computing units. And I think there’s about 15, 15 continuous members. As we speak.
Jonathan Dursi: Very nice. So in all these groups you see and work with you’ve mentioned to me before that some of the peer teams you work with do have some kind of cost recovery model.
How does that work. Are there things you wish you could do cost recovery for, or are you pretty happy with given the funding environment?
Matthew Smith: You know, units that we work with on campus that, that do engage in cost recovery mode. And although I engage with them less now in my capacity to ARC they’re still around in supporting our operations.
But I certainly work closely with them when I was with Psychology it, which is UBC IT Audio. Yeah. So the AV unit, they are completely cost recovery. They need to ensure that, you know, their costs are, are recouped somehow. I believe that might not be entirely, like I’m sure they have some operational funding to, to ensure that they, you know, have, have a baseline.
And you know, just like if you were to go to. Let’s say an AV supplier, if you had a need for, let’s say within research, a behavior monitoring room, you would go to, you know, previous to UBC IT AV being available. You would go to this commercial supplier, they would come, they would run a consultation, they would spec out your room and then basically perform the installation and.
UBC it used to kind of help lays with these, these different individual kind of commercial providers and provide kind of that liaison service and then realize that, hey this is an area where we can bring a lot of value by offering it directly. Purely because every time you have an outside group coming in, they’re spending a lot of time just learning about the infrastructure, learning about how the buildings are laid out, what the network is like what they can and can’t use, et cetera. So yeah, so UBC IT AV, they’ve always been a presence, but certainly in the last seven years they’ve really turned into the folks that you want to go to when performing any sort of AV installation.
So by the very nature, you know, UBC’s huge, as I mentioned, huge community there. So the cost savings are there. Like we, if I need a piece of AV equipment or like a, an AV controller, I know that I can get it cheaper from UBC AV usually then I can get from another supplier unless they’re trying to like, win me over and give me a sweetheart deal.
It just kind of makes sense in terms of suppliers. Also, they know, you know, the needs of UBC so they can bring in, certainly during the supply chain management issues we had with the pandemic you know, we saw them kind of stock up on equipment that they knew the community was was gonna need.
And then I think another you know, a key benefit that certainly my bias looks at is that ability to communicate with the research computing units. Typical AV is, you know, a lecture theater, et cetera. It’s straightforward in that regard. But we are seeing a rise in these needs for these behavioral monitoring rooms, these debriefing rooms where let’s say you have a baby researcher that has, you know, a room that kind of looks like a daycare, but it’s a small one, and they need to have cameras and mics placed in strategically to pick up on, Yeah, on audio. Could be really subtle you know, syllables that they’re looking to tease out how the baby’s pronouncing words or they want to be able to track their eye movements, et cetera.
I don’t know that there are too many units that provide research-specific services. There are a few, but they’re based in the states and they’re gonna be really expensive for you to bring them in. So, you know, the, the second best choice into bringing in these behavioral monitoring AV experts is just to work with your departmental research support guys that do know your research. Your departmental AV team are already being paid by your department typically. And then you have the cost recovery mechanisms to, to bring in, you know, the latest and greatest technology that, that you need to implement. Yeah.
Jonathan Dursi: Terrific. You mentioned that there are so few groups that focus on researcher needs.
Do you know, just outta curiosity, if the UBC AV group provides services, maybe only even remote remote consulting services for other institutions who have similar needs?
Matthew Smith: Hmm. I don’t know that they, they, they would much like art services, all of UBC’s services tend to focus on those eligible UBC researchers or, you know, UBC IT AV, they even deal with the administrative units, et cetera.
So while you know, if there’s a collaboration, let’s say, based in, I know that there are a few of the farm sites that are based in central BC that, that probably involve collaboration from other universities. They’re certainly engaged there, but I don’t know that they would engage directly with an outside unit without that, without some sort of prof. coming from within UBC.
I think there’s, there’s a couple of deviations there that I just wanna make note of. Is we do have a fewer audio visual experts that work within the Alliance capacity.
Jonathan Dursi: Oh, interesting.
Matthew Smith: And they, and I do believe they provide specialized services as needed, though, with their alliance hats on.
Jonathan Dursi: That’s interesting. I’d not heard of that.
We’ve already talked about the team pulling together during the during the pandemic. I had the same experience by the way. I was leading a team that was at three institutions and us all going remote all of a sudden sort of broke down the silos. It always felt sort of like three teams collaborating and then, and switching to all remote really brought the team together.
And you gently tapping on the brakes so that the team didn’t burn itself out when, when the workload hugely increased. What are some of the other successes you’d say you’ve had with this team as a manager?
Matthew Smith: It’s everything. I don’t know. We, we’ve been able to, to accomplish quite a bit. You know, when I first started art, we had the HPC offering our, our sockeye cluster had just gotten off the ground. Only, you know, the, the HPC components were all connected and working together, but we essentially had to move from, we had phase one, we had about half of our nodes ready to launch, and then phase two was just kind of had just been installed right when the pandemic hit.
And in addition our object storage wasn’t fully installed yet, so we had to coordinate that in the middle of the pandemic. And I think the team working together to understand, you know, basically work on each other’s comfort levels in terms of, you know, doing onsite work at a time where you were told to basically stay home or else be stricken by plague.
So just taking, you know, there was a bit of risk taking. The team all supported each other in, in doing, in order to kind of keep the spice flowing while also recognizing that we did need to make more stringent judgment calls on what was necessary and urgent and what was you know, what, what could wait until we had a little bit more freedom.
So I think working together to, to basically ensure that everything continued developing as planned. Like it wasn’t just operations, we had to, to move projects along. Yeah, we, we launched our object storage service using Globus which is a great software based data transfer technology if folks aren’t familiar.
We launched that in the middle of the pandemic. We launched our cloud pilots and and service, which is our Ronin.cloud offering. Cool. Which we call UBC ARC Ronin. We launched that in the middle of the pandemic. Yeah, I could. Yeah, the team really was able to, to accomplish a lot. And I think that came from remaining engaged with the research community, not getting too much in our own heads listening to what their needs were.
And then working to ensure that, that what we spent our time on had the researchers in mind. We didn’t take that, build it and they will come attitude, right? Like we, we ensured that, that we knew who was coming to the table. Like who would take advantage of our services and had them in mind as we were developing them.
So, yeah, I think think that, and then grants that’s something that I, I can’t gloss over. I have to be sure that I, I catch is. One of the, the, maybe the advantages of the pandemic is you did see a lot of these covid 19 additional funding opportunities become available. And my team particularly Jacob Boschee worked hard to secure a lot of support through one of those grants.
So we worked with AMD in their, their Covid 19 fund, and we secured over a million dollars worth of access to equipment for one of our. Our key he, he did social covid 19 research Nice. That then involved into Arabic language processing. His name’s Dr. Muhammad Abdul-Mageed. They just launched the, this toolkit they called TURJUMAN, which is a Arabic language processing toolkit that they attributed access to those AMD resources as.
A key to their ability to develop that. And yeah, so I think that’s something that we’re proud of is our ability to kind of take advantage of those additional funding opportunities while dealing with the, the stresses of right, of yeah, increased demand in operations. And then just you know, dealing with our own responses to the pandemic.
Jonathan Dursi: I know for research computing and data managers, it can be a little uncomfortable to toot our own horn, but what are some of the things you’ve personally done as a manager that you think have helped contribute to the success of your team?
Matthew Smith: I think diving into the coaching based approach to leadership has been of value in the midst of the pandemic. Coaching has always been something I’ve been interested in. But I enrolled in UBC’s organizational coaching program in the middle of the pandemic to, to be able to provide that kind of, you know, non expertise based - I don’t wanna call it mentorship, because it very much is coaching, but basically being able to have powerful conversations with team members without being an expert in what they’re trying to troubleshoot or accomplish.
Because yeah, it’d be a rare bet where we’re gonna have circumstances where I’m the technical expert, amongst the rest of my team who collectively have over a hundred years experience. Coaching is a skill, and that’s where I think I’ve been able to, to bring value to the team is providing space to bring their problems, to focus on a problem or solution they wanna reach and then offer a chance for them to maybe see my perspective or even look at other, other perspectives on, on how they might approach a problem, whether it’s technical or interpersonal.
Jonathan Dursi: It’s a tough transition, at first, for those of us especially who ame up in research. We progressed by being becoming more and more expert. And then to switch to this… Are there resources on coaching that you’d recommend to members of our community?
Matthew Smith: That’s a great question. I’m spoiled a little bit in that I’m part of a, a coaching at, at UBC community, which unfortunately, unfortunately I do think is, is limited to, to members of the UBC community.
Yeah, there’s you know, one of my mentor coaches Andrea Fruhling has a site and blog called Double Knot Coaching and Consulting. And yeah, Andrea’s amazing and certainly was one of the key folks I had an eye on when getting into coaching. So that’s just one example that I can think of off the top of my head.
Jonathan Dursi: Terrific.
And we’ve mentioned not having enough resources or time for all the, all the excellent research that you could be supporting. Are there other challenges? You face right now as, as the manager that, that you’re wrestling with that, that some of the community members might, might sympathize with?
Matthew Smith: Hmm. I think understanding the changing landscape in terms of funding, collaboration, et cetera. Certainly up here in Canada we’re facing a little bit of a shift with regards to how federal resources. Are shared and distributed to our researchers and trying to stay on top of that ensure that we’re involving ourselves in the conversations we need to, to champion our research community’s needs while also not getting in the way of developments is, is certainly a challenge right now.
You know, certainly since I’ve been here at UBC most of. Federal and regional services were delivered by either West Grid or Compute Canada. That’s all been taken over now by a group known as the Alliance which are, are just starting to, to release their vision for research computing in Canada.
Right. So I think staying privy to that while also, you know, and, and being able to respond to it while also providing a firm vision for what we are going to be working on. Right. It has certainly been a challenge.
Jonathan Dursi: Terrific. And are there any final thoughts you’d like to leave their research computing teams community with?
Matthew Smith: Just, just love what you do, love the people that you support. I think all of us in research computing are in a unique position. You know, it’s quite the niche that we’re involved in here, and I think we have a rare opportunity to contribute to what I certainly see as one of the key parts of society, which is, is research, new learnings, new innovation, and being able to support that is a rare opportunity. So I would say if you’re in this opportunity, make the most of it. Put all you can into it and the world will benefit.
Jonathan Dursi: That’s terrific. That’s great. It was fantastic speaking with you, Matthew.
Matthew Smith: Thanks so much. Yeah, likewise Jonathan. I really appreciate this opportunity and thanks for having me on.
