Moneyball for Software Engineering | Jeffrey Haynie from Pinpoint

Artificial Intelligence is helping us work smarter and faster in every industry from clothing to cars. But can artificial intelligence help us build better artificial intelligence?

In this episode of UpTech Report, I interviewed Jeffrey Haynie, the CEO and co-founder of Pinpoint, which uses AI to help software developers gain better insights to provide efficiency and deliver stronger results. Pinpoint even uses their product to help improve their own product. Jeffrey discusses his efforts to build a common language to better communicate the process of software development and the challenges of being at the forefront of his field.

Inside of the software factory

What’s happening inside the software engineering process? Only engineers can understand that? The idea of Pinpoint is to try to help software leaders to better understand it.

Jeffrey himself is an engineer and they try to uncover what’s going on inside of the software factory.

“We think of a lot of software as a pipeline. If you thought about software, at the end of the day, ideas sort of go into the top of the pipeline and sort of comes out, as a working product, free from defects, doing all the fun, and hopefully achieving the business outcomes”, says Jeffrey.

The problem is what happens in the middle of the pipeline, something that usually non-technical people can’t understand.  Here’s where Pinpoint comes in.

Jeff is co-founder and CEO at Pinpoint, helping companies unlock engineering performance to build software better. Previously, Jeff co-founded and scaled Appcelerator (home to Titanium, a development toolkit used by more than a million developers around the world) to acquisition by Axway in 2016.

Jeff is a long-time serial entrepreneur and angel investor—prior to Appcelerator, Jeff was Co-founder and CTO of Vocalocity, and CTO of digital incubator eHatchery. He has worked on numerous standard committees from IETF to W3C, and contributed to several open source technologies including JBoss and OpenVXI. At the beginning of his career, Jeff served with distinction in the U.S. Navy.

DISCLAIMER: Below is an AI generated transcript. There could be a few typos but it should be at least 90% accurate. Watch video or listen to the podcast for the full experience!

Alexander Ferguson 0:00
Artificial Intelligence is helping us work smarter and faster in every industry from clothing to cars. But can artificial intelligence help us build better artificial intelligence? In this episode of UpTech Report I interviewed Jeffrey Haynie, the CEO and co founder of Pinpoint a company that uses AI to help software developers gain better insights, improve efficiency and deliver stronger results. And pinpoint even uses their product to help improve their own product. Jeffrey discusses his efforts to build a common language to better communicate the process of software development, and the challenges of being at the forefront of his field. Thanks, Jeff, for joining us, I’m excited to learn more about Pinpoint kind of the speciality in the area that you guys are focusing on and using your technology make the world a better place and you as a entrepreneur as a, as a leader, how you innovating? How are you growing? To start us off? Tell me, where did you start? What do you what year did pinpoint start? And where are you guys located?

Jeffrey Haynie 0:59
Just a few years ago, and we’re currently headquartered in Austin, Texas, although we have people in different places.

Alexander Ferguson 1:07
Are you bolter bootstrapped or VC funded?

Jeffrey Haynie 1:11
Or venture capital funded?

Alexander Ferguson 1:13
Enterprise? Okay, how big is your team now?

Jeffrey Haynie 1:16
We’re a little over 35 ish right now. So not too terribly big. Okay. Small,

Alexander Ferguson 1:22
small, but but you’re able to do mighty things with the team. Right? That’s right. Tell me more about the industry that you’re you’re playing and kind of the the role that you’re serving in that industry?

Jeffrey Haynie 1:34
Yeah, so we’re trying to go after new category we’re trying to create, if you will, a category we call engineering, performance management, and really trying to help software leaders better understand what’s happening inside the software engineering development process inside there’s companies.

Alexander Ferguson 1:50
So what does it look like? What pain point are you able to then solve for those engineering?

Jeffrey Haynie 1:56
Yeah, so that made the biggest problem as an engineer, me and my co founder started the company, because we’re engineers, longtime engineers, and have started many companies and really struggle with engineering itself, because engineering is such a black box, right? So we’re trying to really help business people, technology leaders really on on, you know, demystify if you will uncover what’s happening inside the Software Factory. And the way we do that is we think about software as a pipeline. If you thought about software, you know, so Id, at the end of the day ideas sort of go into the top of the pipeline, and sort of what comes out, hopefully, is working product, free from defects and doing all the fun, cool features that you’d built for it, hopefully achieving the business outcomes. But the problem is sort of what’s what happens in the in the middle of the pipeline is pretty mysterious, especially to non technical people. And what we’re trying to do is really help illuminate what happens inside that funnel, or inside that software pipeline. And then by doing that, we can actually create, if you will, a common way that people talk about software as a business and help them understand common, if you will, common vocabulary about you know, how the software as a business operation works,

Alexander Ferguson 3:07
what does the package or pricing and then the end solution for that? What does that look like for them? Yeah, I

Jeffrey Haynie 3:13
mean, it’s a typical SaaS product, basically. So it’s licensed on a SAS annual subscription basis, delivered over the cloud, like most SaaS products, and is priced on a on a per user basis, the number of users that we track data again, so not the end users that use our product, but the number of sort of, if you will, data subjects or people that we track the data against, that’s how we license it.

Alexander Ferguson 3:38
How many customers do you have right now, on your platform, we

Jeffrey Haynie 3:41
have a handful of customers, few big customers, we’re still in pretty early. And we were in private alpha for a couple years, and with some paying big customers that we you know, one of the challenges of building this type of company is a data analytics company. And you need real data, especially for building our machine learning. So we spent the first couple of years really just invested in a handful of companies that we could get their data, work with them, kind of refine the product. And over this past year, we really started to add new customers and really looking to grow grow, if you will, the external opportunity

Alexander Ferguson 4:12
for one of your average companies, what’s like the number of users that you’re end up tracking data for?

Jeffrey Haynie 4:20
Usually, it’s in the you know, we tried to go after, you know, a broader segment. So today, it’s usually in the multi 100 size developers that they would have a lot of our new growth and new pipeline that we’re adding a lot of interest is in smaller company. So companies all the way down to 2030. Developers, we found that it’s been quite quite a good product, we use pin point for ourselves, we’d like to say we’re customer zero. And so you know, we’re 20 ish engineers and data scientists. And so we’ve been trying to eat our own dog food and trying to build the product for ourselves first, with this idea that if we can use it ourselves and sort of you know if it gets scale, if you will, down to even relatively small teams, then you can certainly scale it up to much, much larger companies. As you get bigger. The problems, of course, get magnified, they become different in some ways, but the core infrastructure, the core algorithms, Core Data, things like that are pretty much the same, whether it’s a 10 person development company, or 10,000,

Alexander Ferguson 5:19
if you had to give an analogy to what this does, for someone who’s non technical, what would you how would you explain that?

Jeffrey Haynie 5:27
I mean, I think the easiest analogy, especially for business people is just Moneyball for software. If you if you sort of see what happened with baseball, and professional sports in general, you know, when analytics came onto the scene, you know, many years ago now, but when analytics came onto the scene, and we started really leveraging analytics for how we actually built high performance, you know, teams, and how you pick players, and how you sort of figured out sort of how to track and where to where to sort of play the game differently, because analytics, I think, you know, engineering is probably a good analogy to that. If you just sort of use a, you know, sort of sports analogy,

Alexander Ferguson 6:03
if I read correctly on your site, you’re, it’s supposed to be very easy on the actual company that puts it in, it should just connect to the existing platforms and tools they’re using and just takes tracks the data there, did that capture capture that right? That’s right.

Jeffrey Haynie 6:19
I mean, you could think of it like telemetry, so we sort of integrate into your existing engineering stack. So if it’s a GitHub or git lab, or if it’s Jira, or if it’s a, you know, you know, whatever, your CI, CD tool, etc. And so we the first on board is we basically, once you give us permission, via usually the O auth. Or if you’re on premise, you might use an API token or some sort of share key. Once we get access to your data, we basically go back to all you know, all time, so we look at all your historical data, using the publicly available API’s for those products. We pull all the data into our system. And then of course, we we we sort of have a real time pipeline that then processes that data, you know, uses machine learning models, things like that, to then create basically, ultimately, an application that the customer interacts with, mainly today via the web. So a desktop application that they would SAS to cloud deliver an application that they would log into. And so what’s nice is, it’s sort of you know, from day one, you can get immediate visibility and capability and value. You don’t have to go spend six months deploying it and changing behavior and doing a bunch of things to be able to get value out of it, we can look immediately at what’s been happening over the last two or three or four or 10 years. And you can see through pinpoint, and then of course through that, then you can start to actually figure out where do you want to invest your energy? And maybe where are things you want to change or improve and start to use pinpoint to really help you get that.

Alexander Ferguson 7:52
The Machine Learning then which you said, the first kind of year, so you were focusing on and fine tuning? Has that been an interesting process? How has that developed now? Because that’s what allows you the results?

Jeffrey Haynie 8:08
Yeah, it’s interesting. So you know, there’s sort of good news and bad news for us on I would say the machine learning data science sides on the on the sort of the good news is that these problems, we’re going after somewhat toy problems today, and artificial intelligence, machine learning, meaning we have highly structured data, if you think about it, you know, a JIRA ticket, if you use a simple example has a lot of interesting labeled information, it’s got priority, and owner and team and due date, and all kinds of things, probably things like components that may be related, maybe it has links to other issues. So there’s, there’s a whole graph of data that’s highly tightly typed, and it’s not. And it’s highly structured. So now, there’s usually no correlation between that thing and all the other data that sort of exists that we pull in. So that’s where sort of starts to get really interesting. The second part is the sort of harder part maybe the opportunity for us, but certainly the challenge from a, from an engineering standpoint, is that, you know, there is no, there’s not really a huge body of technology that exists out there to do what we do. It’s not like, you know, self driving car, and there’s a whole bunch of NIST data sets or things like that, you can just pull off the shelf with a bunch of algorithms that you just really need to tune and do feature engineering and things like that. So we’ve really had to actually create you know, we’ve had to sort of create an invent a lot of things like how do you analyze source code? How do you extract you know, sort of meaning out of get blamed data, things like that, this sort of build a model, and then a lot of sort of iteration around, you know, what is interesting? What can be used to actually do predictions or forecasting what can be done to, to look at things like code risk, you know, how do we sort of take all this, you know, sort of unrelated data from a data model standpoint and create relations and correlations and try to figure out causality and things like that and so on. It’s it’s a, it’s a very interesting scientific and engineering problem, because, you know, we’re sort of inventing how to do these things. And and the only way you can do that is just a huge amount of iteration,

Alexander Ferguson 10:12
any learnings that you can share on an overall basis of growing now, actually, two companies have insights of hurdles you had to overcome, that other entrepreneurs and leaders can learn from?

Jeffrey Haynie 10:26
Yeah, you know, I guess I got a lot of lessons learned. I mean, you know, typically entrepreneurs, you just make tons of mistakes, even even when you know, what you’re doing, you know, and I guess that’s part of the journey is right, you know, there’s no roadmap to what we’re doing. I mean, there’s a, there’s a sort, I would say, a general thesis about how you build startups and how you raise money and sort of how you build sort of the, you know, sort of get to market with an MVP and things like that, of course, any company does that. But sort of in the sausage making, there’s all kinds of things that happen that just aren’t obvious. And so, you know, partly what we’ve just tried to do is, you know, you hire the best people that you possibly can, you know, sort of empower them to understand what the mission is, and what the vision of the company is, and kind of what we’re trying to achieve, and try to find sort of customers early enough that, that sort of buy into that vision and early adopters, in general, that will work with you. And we’re sort of worked through through it with you to sort of become design partners. And then we just do a lot of iteration, right, you just sort of have to like iterate like crazy, because, you know, partly, you’re trying to get product market fit. You’re trying to sort of build a culture and a company around sort of this idea, and ultimately trying to sort of figure out how do I build something profitable, that can scale that creates value for my end customer. And then we can actually, you know, build something long term and sustainable and that, that just, that doesn’t happen, because you read a book or, you know, you sort of go off and do exactly what you said you were going to do. Because you know, what you think about doing in the very beginning, and how it sort of ebbs and flows and how it evolves as you execute, you’ve got to be really sort of dynamic, and that and really kind of keep your ear to the ground. And so, yeah, that’s sort of what you know, my, you know, I would say, operation has been for the last, you know, several venture backed companies I built and, and, you know, success is never guaranteed. So you just try to do the best you can to listen to the customer, listen to your culture, as a company and try to build the best thing you can.

Alexander Ferguson 12:21
Looking forward from here, you have a vision of the direction, what where do you see pinpoint in five years from now,

Jeffrey Haynie 12:29
you know, in five years, I mean, you know, who knows, right? So for me, it’s like, I think five years, it could be a substantial company, I think we can create a category that a lot of people use our software, we think we create a platform and an ecosystem around that software that you know, that that allows a, you know, not just what kind of innovation we can do in our own product, but you know, sort of an ecosystem of partners and people around us that can help us create value around our product, if it’s going to be a very large, which we think it’s going to be a multi 100 billion dollar opportunity long term, because this is a big category. You know, like I said, we’ve got a vision and a roadmap, if you will, but it’s sort of hard to know, when one of the things that’s hard about building a new category, if you’re building something better than it already exists, right, you’ve got a very good understanding of the market, the the budgets, maybe that customers have kind of the limitations that the market might have, or products that are out there. When you’re building a new category. It’s literally Greenfield, right? So one of the awesome opportunities is that, because it’s Greenfield, you can sort of design what you want, and sort of sort of build what you what you vision into and sort of get, you know, sort of market feedback. But also customers don’t know what they don’t know, they don’t know what they want, and they don’t know what they need. So there’s, it’s, it’s very much a missionary evangelicals sort of early opportunity where you’re trying to help them, but they need to also help you, right, if you ask them exactly what they need, in, you go build that you won’t, you won’t build a new category, right? You’ll, you’ll sort of prevail the proverbial faster, you know, you’ll, you know, so, you know, for us, it’s a little bit like, you know, really trying to sort of work to solve, really, you know, we want to be, you know, an aspirin, not a vitamin, we really want to solve real problems. But we also have to recognize that today, people don’t have this. So, you know, we need to also educate the market, help our customers understand the value that we know they will get, and also sort of help them if you will, sort of bring this into their organization. And then as they mature themselves, and as the overall market matures, then of course, things will actually continue to you know, sort of evolve much more rapidly.

Alexander Ferguson 14:34
You’re having to take on the ability to innovate and think of new ideas, how do we structure this? How do you innovate? What do you where do you look for for new ideas and new ways and thinking of solving

Jeffrey Haynie 14:45
problems? Yeah, so you know, probably our biggest challenges we have no shortage as entrepreneurs and technical entrepreneurs and and ultimately, sort of the end user of a product like this sort of North no shortage of, of probably good ideas, and that’s probably the hardest part we we have Really trying to distill, like what is really truly a good idea and really innovative, because we can sort of virtually do anything. And probably we should, we should virtually do very few things. And we should do very few things, but do it really, really good. And that’s probably the hardest part of any startup, right, we could, we could do virtually anything, we have sort of this unlimited supply of ideas and the data to sort of do interesting things. So trying to find, if you will, the sort of threaten the needle between you know, what you can do and what you should do, and where the highest value that really solve real problems to the customer. And what would sort of be a nice to have, and maybe isn’t really super, you know, super valuable. You have to try a lot of things, which we do get a lot of feedback. And then you have to be willing to iterate and sometimes throw things away or start over from scratch. And that’s kind of how we do it. Me and my co founder had been pretty good at doing this for a long time. You know, it’s our third company together, you know, we’ve been doing a lot of time. And so, you know, I think our willingness to sort of, you know, be creative, but also sort of listen to the sort of customer what the customer is asking for, not necessarily what they’re saying they want, but sort of what their really problems are, what they’re trying to solve for, and then trying to build something towards that. And that’s kind of how we do innovation.

Alexander Ferguson 16:19
Any podcasts or books or audio books that you’re reading or listening to right now.

Jeffrey Haynie 16:24
Yeah, in podcast wise, I’ve been I pretty much my daily podcast is software engineering daily, which is a pretty popular podcast, Jeff does and and, you know, it’s got a pretty good range of topics more technical. And sometimes, you know, there are some business or product oriented conversations, but often they’re more technically inclined or new products or new ideas and things like that. So that’s, that’s one vehicle for me to kind of keep in touch with what’s happening and new information that’s happening kind of real time. I do listen, shamefully, I was to, you know, Andreessen Horowitz is they have a few that have pretty good podcasts. And they have some pretty good content, especially for startup oriented entrepreneurs that are thinking about how do you build scale companies? And so I’d say those are probably the two biggest things that I’m that I’m focused on from a podcast. I do listen to a ton of podcasts. Those are the things that I listen to on a podcast basis.

Alexander Ferguson 17:18
Thank you so much for sharing all this great insight. Where can people go to learn more, and what would you recommend is the first step for them?

Jeffrey Haynie 17:24
Yeah, it’s easy, just go to pinpoint calm. And, you know, first step would be you know, read some more content, give us feedback. Tell us what you’re thinking about. You know, we’d love to help. And I’m easy to find on Twitter and pretty much anywhere else. So

Alexander Ferguson 17:38
that concludes the audio version of this episode. To see the original and more visit our UpTech Report YouTube channel. If you know a tech company, we should interview, you can nominate them at UpTech Or if you just prefer to listen, make sure you’re subscribed to this series on Apple podcasts, Spotify or your favorite podcasting app.


YouTube | LinkedIn | Twitter| Podcast

From Boardrooms to Bistros | Alison Galix from Dinova

Business Lessons from Tolkien and Stephenson | Dipanwita Das from Sorcera