Talent Development in Open Source with The Linux Foundation’s Clyde Seepersad at OSS Seattle 2024

This is Textron tv. Hey everybody. We are here at the Open Source Summit in Seattle, Washington, talking with many, many great people.

One of the great privileges we get to spend some time with the Lenon Foundation folks from different parts, uh, different on the foundations within it. We're gonna talk about learning and education and training and hiring. And sometimes every leader, I dunno, struggles with, right?

How do you get the people that you need? Not only now, but uh, somebody you need 12 months from now. You can help get them there and help them advance in their career.

So I'm joined by Clyde. Super sad. You nailed it.

Thank you. Good pleasure To be here. Thanks, sir.

I'm glad I got that right. Um, tell us, tell us about what you do at the Links Foundation and tell us about yourself. Sure.

I, I've been at the, the foundation for a decade now, working in the one area that I think most people don't realize we work in, because when folks hear about the foundation, they think about the 800 plus software projects. Absolutely. And the massive developer communities and the ecosystems of usage.

While it turns out you don't just learn this stuff spontaneously, somebody's gotta help you onboard. And one of the challenges with open source is because it's not coming primarily from a vendor who's writing the entry level training material, who's developing those entry level certifications? Are they high quality, affordable, globally available?

And that's what this group does, is we focus on the projects are great, we love the developer communities. How do we make more people a aware of what these things are and the implications for their business? And B, technically competent to use those, uh, projects within their operations.

Uh, so they can actually then become part of the user community and get that nice self ling model of growth. But so it's training's an important piece of it. It Absolutely is.

So it's, it's about how do you use open source software in your, in your work, your organization. You also do training about how to be a good open source contributor Absolute thing. Okay.

Because that, if you haven't done it before, I mean, you'll learn from others. There's An SBO and why should I do it? Yeah.

How do I think about the different type License? Get that the drug store, is that the candy? Where do I get my S bomb best practices for?

Uh, yeah. You don't order it all on details. Right.

Uh, how do I think about secure coding practices? Right? Mm-Hmm.

You know, we do a whole spectrum of stuff. Developers as well as users. Great.

Is there different things for different parts of Linux Foundation, or do you like, um, you know, continuous delivery versus another part? Or is it just apply to everybody? So we work as a centralized organization and all the, we work with all the projects.

And what's been true in pretty much every case as we go to our project and say, what are you doing on training? And they look at their toes and mutter about the Wiki, uh, because they're passionate developers, that's not their thing, right? Yeah.

They think is the right great code. And so we provide that centralized service that says, Hey, we can help you with the design of it. We can help you with the engineering of the delivery, with the selling of it, with the customer support desk.

That's 24 7. The piece to run, figure out whether it's e alerting, whether it's instructor led. And so we really provide, when it's a plugin service for all the things that you need to do to have a talent development program that go beyond having subject matter expertise.

Mm-Hmm. You know, in all honesty, you know, I've run engineering teams before too. And I I didn't know that the Linux Foundation did this.

We have to help people. You, which is why we're doing this. This is Why we're doing this.

It is, to be fair, it's not something we've always done. Okay. Is a relatively new Five years old.

It's 10 years old. Okay. We've been, we've been at it for a little while.

Yeah. Clearly we have more work to do to raise awareness of the fact that just like we have an events team that puts on this amazing event that we're at, we, we do have a talent development team that tries to work with developers, with individuals, with organizations. Say, let us help you get up to speed to make the most out of these projects in your operation and to contribute to these projects.

Don't just be a user, be a contributor. Yeah, absolutely. It's part of creating a vehicle ecosystems.

That is. It is. And a lot of people wanna know how to get involved and you know, it, it's a little intimidating, right.

It can be a little intimidating. There's some very talented people working in these projects and you don't wanna be the dumb one, but, you know, you gotta Stop something I've figured out long ago. I'm never the smartest one in the room.

So just don't worry about it in this room In particular. It's especially in this room. Absolutely.

So let's talk about, um, a couple things. One, we were talking earlier about moving to the cloud, right. You know, startups and greenfield kind of things just to kind of pick your provider and you can begin and go a little tougher, but a lot more difficult for enterprises and people that have a legacy or more complicated, uh, applications and infrastructure.

It is, you know, I think the analogy I've been using is that, uh, large organizations have been at this for more than five years. When they get onto the therapist couch, the first thing they confess is there's a lot of stuff that's not on the cloud. There's a lot of legacy stuff.

Oh, yeah. Yeah. There's a lot of stuff running on mainframes and things that are running in, in, uh, architecture that maybe only a dozen people in their organization know anything about.

And their worry is that it's very sexy to talk about what you're doing with your app and your cloud and the stuff that you've moved and how you're integrating ai. Mm-Hmm. It is really not sexy to say, I'm gonna refactor all the stuff that works fine today, but is running on hypervisors and mainframes and physical servers in our co.

But it has to get done. And so I think this recognition that we've gone through the initial phase of falling in love with the cloud and all the benefits it brings with scalability and regionalization and low latency and costs, so long as you manage the cost. But now we have to pull back and think about the reality of, there's a lot of it that's not on the cloud, and how are we gonna get there?

How are we gonna move that legacy infrastructure without spending tens of millions of dollars and, you know, five years on a moonshot program? Mm-Hmm. A because nobody's gonna give you the money to do it, and b, because it's just not kinda where you want to be focused when you've got some of the new business stuff.

And so I think part of what we're trying to, to have more discussions about is helping organizations figure out where, in a world where your reality is very much heterogeneous computing Mm-Hmm. What are you gonna do for talent? How are you gonna get from where you are today to where you need to be leveraging as much as possible the talent that you have today?

Yeah. I don't remember any cost analysis spreadsheet saying, check this if it's the right thing to do. Right.

That's your justification. It is. You know, and I've learned that lesson long ago.

Like, we need to convert to this data. Well why? What's the value?

You know, one, one of the things, my experience going to the cloud at least one, one round was use of capital that I didn't have the capital to, to refresh. And I could have gotten it, but then I'd be taking it away from bonus and using that kind of thing when you're running kind of a cash flow type business. Um, and that made it much easier to justify why we were gonna do this.

Um, 'cause it, it did save us money in at least in that period of time. Is that a very common, you know, justification or less so it's more complicated than that. You know, as with it's always complicated.

It's always more complicated. I think you could safely assume that everybody moved the stuff that was easy to move. Okay.

Over the past, Well, hanging fruits done. Okay. Right.

Yeah. It's the stuff that wasn't easy to move. It was a step that's deeply entwined with your back office workflow processes that you have limited knowledge base even within your organization.

'cause stuff was built in the nineties. Uh, that's a system that hasn't been, I think we've gotten through that first wave of, if it was easy, I'd rather do that than go and ask for $5 million in node balances and barracuda servers and actual servers and storage because I can move it to an opex cost. Right.

And that was, that first wave was, hey, move CapEx to opex. Your CFO's very, very happy. Yeah.

And of course even that on the backend now, the whole finops practices because you know, suddenly you're getting Yes, We know a lot more now A seven figure bill. And you're like, well is this actually cheaper? 'cause I'm paying it every month.

So there's a whole evolution on that side. Mm-Hmm. But what's been true, I think is there is a lot of stuff that wasn't candidates for an easy lift and shift and something's gonna have to happen.

Mm-Hmm. And how are we gonna get over the hump of the stuff that was hard that we sort of didn't do in phase one? Yeah.

It was interesting, right? A few years ago when, uh, when the sub microsystems licenses finally said, okay, no more support and organizations came out of woodwork. What do you mean it's been 20 something years?

They've letting you know for quite some time. It's end the life. It was Well communicated.

Right. Um, that's not where you want to be. Right.

You do not want to be to a place where you're up against the wall critical production system is literally no longer supported. That's gonna get expensive and it's gonna get painful. So from a risk management perspective, figuring out ways to chip away at that stuff without a whole ton of capital is where I imagine most organizations, it's been there To, that is not a fun place to be.

You know? 'cause you're, you're telling your, your leadership, whoever, whoever's the decision makers are a part of that, of here's the risk, this is what will happen. And hopefully you don't have the big thing, the big problem happen.

Well this is the irony, right? Yeah. You're a tech leader, the defense of, I've been begging you for years and you've been telling me why would I spend money on it?

Yep. And I've been saying it's because the risk is building up. And now the chickens have come into That's Well, and you say it's on you.

Yeah. I told you so. It Always does.

I told you so is not a great strategy. Yeah. I told you so That's not, not another box on that, uh, value proposition spreadsheet.

So, so what, there's technical skills, right? That you want to train people to migrate applications and technology into the cloud. Are there other things that, you know, there's just the fear factor.

There's the risk. You know, I don't wanna put my career on the line and I like what I'm doing now. And you know, that's, that's also part of this human element.

It is very much true. I mean, even for the basic stuff, there are a lot of developers who built a lot of the systems that are running a lot of the computing route Mm-Hmm. That grew up their entire career on waterfall development models.

Mm-Hmm. And so when you talk to those folks about DevOps their eyes to be Uber, It's like starting over. Why I'm good at The way I do it.

It's not that they couldn't understand the technical tools. So they don't understand the cultural shift that goes from, Hey, I'm a developer. I'm writing code.

I'm putting it into a pipeline. It's live today. Mm-Hmm.

And what are the implications? Oh, that's very, yeah. That is a change management piece.

It's a big, not just a technical, I mean it's also a technical skills issue, but it's bigger than that. Right. And so when you're thinking about talent in your organization, you have to think about both channels.

You have to think on the change management application implications from this different way of working. When you get into microservices and you have to take, think about the technical skills piece, which in a way is almost the easier piece. 'cause all technologists are geeks and they sort of build stuff all the time.

But you have to do both. And most importantly, you have to recognize that this is not a 90 day sprint. You know, this is a multi quarter, possibly multi-year effort to sort of chip away at the stuff, build out the skill sets that you need while making sure the plane stays in the air.

Right? Mm-Hmm. These are production systems that we're talking about.

And so slow and steady, how am I gonna get there? What's my strategy for the people I'm gonna be dependent on? And how do I stay patient with that strategy over the next year or two so that we can gradually move our way out from those fantastically stable legacy systems?

How do we get there together? Not, you know, collateral damage along the way. Right?

Absolutely. Yeah. Let's, let's talk a little bit about the hiring of, you know, nobody there, there's never enough technical talent.

Like, we're always looking for more talent, whether it's security or cloud or whatever it might be. And, uh, I just laugh at laugh when I see managers hiring, you know, put out the job descriptions that's got eight technologies, have to have 10 years experience and half of 'em are three years old. Right.

There's no people, like, there's no people on the planet that have all eight of those. And if you really just draws out that hiring process for me, I, I learned, I want transferable skills. I want people that like to learn.

If, if you can do that, then we can take your ruby and your Python skills and you can learn go, or whatever it is that we're doing. Right. Because we're gonna need those skills too.

Um, but it isn't, uh, you know, I have three round pegs. I'm looking for three people, three places to put somebody. Look, it's human nature.

Every hiring manager wants to hire themselves from five years. That's, that's a good way to look At it. Plug them in.

Every hiring manager knows from paying experience those people don't exist. Mm-Hmm. We recently at this event released our 2024 tech talent report.

2 months is the average time it takes to interview, recruit, hire and onboard. So, wow. When you're bringing in external talent and you're thinking, oh, I will just hire someone.

Yep. 2 months, that's not even the bad news. The bad news is that after all of that, 38% of those people are gone within 12 months.

So you need a pipeline. Yeah. You cannot, you know, I keep telling people all the time, talent is something you have to buy, but also build, build.

You cannot hire your way to the future. Mm-Hmm. There's too many other people.

You know, as fast as you're bringing them in the front, somebody's steaming them out the Ballot. Always be somebody that will pay 'em or nope. We're you have to keep hiring.

Mm-Hmm. No doubt. But you can't only focus on hiring.

Right. And that's also a bit of a culture change for hiring managers Right. To be, to think about their existing employee basis, people who may be, aren't in their organization as part of that potential talent pipeline.

Mm-Hmm. Is there, um, that 10 month period to hire somebody, is there any way to kind of tease that apart about why does it take that long? Yeah, the report has the details.

I wanna say it's something like six months to find. I'm hire them, get them to start on four months of onboarding. Um, again, because the magical person who knows all your stuff didn't actually exist out there, the unicorn, Purple and striped and Show that how it all works.

Right. Um, but anyway, you slice it. It's, it's just a prolonged process when you're trying to bring in external talent.

You know, sometimes the hiring process, the front end is just figuring out by, by starting to look at resumes or talking with people, interviewing people. It's just figuring out what you actually really want. 'cause when someone leaves, you're not gonna replace the same exact talent one for one.

It's a time to kind of rethink about, well maybe we should shift some things around or what do, what do we need now that for thinking? And that's part of why it takes six months. Right?

Yeah. There's a lot of examples of, I opened the rec and then I realize, oh no, and I changed the rec and now I've reopened it. And now I'm trying to figure out, well maybe it's two recs.

Uh, it's it, you know, there's, there's a lot that goes into when you lose critical pieces of your organization figuring out what does that backfill strategy looks like. You know, it's not widgets. Right.

One for one repeat. There's times where it could be Right. But a lot of times you're try, you're learning as you're going through the process.

Seems like the recruiting game, if you wanna call, that has also changed a lot. It is. In the last 10 years, It is astonishingly expensive to hire.

Yeah. Yeah. And the only thing Lewis, is not hiring a recruiter and try to do it yourself and then you're up till midnight screening regime case.

Right. So, Mm-Hmm. It's just, it's a difficult process.

It's an expensive process, it's a painful process. It's a lengthy process. Um, so let's not just do that.

Right. Let's really think about other ways to, to get the talent pool that we need within our organization. You know, if you have a, you know, a happy organization, you're, your, your technical staff is, you know, likes the work they do, they like where they work, then that's a great source of talent also.

Right. They, I know somebody who is really good at that or would like to move into this and Yeah. And this gets sort of the change management question.

Mm-Hmm. There is a a lot of deep seated worry that if I train new people up, they're just going to, Well, I, I'll tell you what, guaranteed a hundred percent they will leave at some boom. So it's, that's a fact.

So why not train them and Why not train? You know, They'll either die in their chair or they'll be, either way they're gonna leave the organization, hopefully don't die in the Chair. That's right.

It depends which way the close appointed. Yeah. Uh, they, they, and what I say to folks is I hear the concern, I want you to think about this.

You're spending maybe between 20 and $40,000 on a recruiter. It cost you maybe $2,000 to upskill someone. Mm-Hmm.

Upskill 20 of them. Right. For one position, for one Recruit, how many $22,000.

Try to see what happens. Yeah. Same 10 months that you were taking to get the seats filled.

You can cross train a batch of folks. Right. This is a financially prudent thing to do.

Mm-Hmm. It's a much lower risk thing to do from a people and fit perspective. 'cause you're taking people from within your, your existing organization.

Mm-Hmm. And if you're so worried about people leaving train a couple more than you need, what you will find is there's better loyalty when you give people paths to grow their career Absolutely. Into something that's really interesting than maybe what you're worrying about.

And so give your folks a chance they might just pleasantly surprise Mm-Hmm. You know, it seems like the other part of that formula is, and, and hand trying to create or have interesting opportunities for them to continue to learn or apply those things from the job. So you're trying to, you know, bring in a new language or start using a new tool or something or moving to the cloud.

Yeah. That's a great thing that, 'cause then they get to apply it and it's not just, I took the training, I did this absolutely this thing, I accomplished This. And you know, one hopeful note we saw in the 2024 tech talent report is that about as many companies said they were gonna invest in skills for Gen ai as said they were gonna try to, to reduce their overall force.

Because I think those exact same wheels are starting to turn to say, well actually I have people in my org that I can trade up on this stuff. Whether it's prompt engineering or thinking about data. And so there's a path to that future that relies on investing back into my existing organization.

Mm-Hmm. Um, so, you know, I think people are beginning, they're coming around, right? As we're getting into this new cycle of wave after wave of technology, the light bulbs are going on about, I, you know, I have these folks who are existing assets who are known entities.

Maybe I should be thinking about being more systematic in terms of how I invest in them to help me get to where I want to be. Yeah. That my thinking is change or be change.

So when something comes along like this is gonna have an impact on us. Let's, not defensively, but kind of offensively or, or, and taking advantage of how could we, let's come up to speed on it, let's learn it so we know how to use it and then we can be at least part of the change or maybe we'll help lead that transformation. You know, the biggest single opportunity from this new generation of LLM models is we had many years there where organizations felt that they had to hire PhD data scientists to get into the game.

Good point. And now they realize, oh, I don't need to have million dollar data scientists on my payroll. Plus many smaller organizations couldn't afford And we're not that many of 'em out there.

And so it's opened up this frontier, right, where you can have people who aren't as deeply technical as what you would've needed even two years ago. Mm-Hmm. Take advantage of it.

Right. I think out much quicker, people can come up to speed. So yeah.

It's insane. All good stuff. Well, in part two of our conversation on that, we could go on for a long time.

I really enjoyed talking with you, Clyde. Thanks. Um, we wish you the best and thanks for, you know, you doing God's work here, training people, helping them in their careers and I as saw you helping us get the word out.

Absolutely. We'd love to have you on some more stuff too. Rip a lot Up.

I really appreciate the chance to get with your audience. You bet. We have a lot of conversations about upskilling, re-skilling, all those kind of things.

Soly, um, sza who is with the Linux Foundation, uh, on the learning, on the training on of it. So, uh, you know, invest in your people. Phil, appreciate you even more.

So we'll be back with, uh, some more great interviews. Thank you Goly. Thank you.