Treeline CEO Peter Doyle Wants His MSP To Use Agentic AI ‘To Fundamentally Change This Industry’

‘We’re setting out to try to build what we perceive to be the modern or next generation service provider model,’ Treeline CEO Peter Doyle says.

Treeline CEO Peter Doyle sees the AI-era MSP he’s building as a way to overcome the solution provider struggle of adopting the latest emerging technology when already low on resources and wary of diluting margins through new software purchases.

“The pace of technology today–and now AI, general automation software–is moving so quickly that we just don’t think that the industry as a whole is necessarily prepared for that or able to keep up with it,” said Doyle, who co-founded Treeline in 2024. “For the first almost year of the company, we’re integrating with these legacy tools. We’re building infrastructure so that we can then leverage AI agents and just good software to fundamentally change this industry. So it is going to be fast paced.”

Helping Doyle on his quest to revolutionize the solution provider market is a $25 million Series A round of funding, with storied venture capital firm Andreessen Horowitz (a16z) leading the round.

[RELATED: AI Inference Software Startup Gimlet Labs Raises $80M In Series A Funding]

Treeline Series A

Joe Schmidt, a partner at Menlo Park, Calif.-based Andreessen, told CRN in an interview that so far he’s seeing Treeline’s AI-era approach succeed through organic growth rates unlocked by capacity increases from the existing team.

Treeline has the potential to realize the MSP dream of calling back all those customers they don’t have time for and more easily providing existing customers more services thanks to time savings and faster product development within the MSP.

“Every MSP, you have so many more customers that you could serve that you’re not serving because there’s no way to actually call them back,” he said. “There’s just a lot of latent demand. The addressable market is literally every small, midsize company in America. That is an enormous market.”

For Treeline to achieve near-term success, Schmidt wants to see the company build out a sufficient geographic distribution of highly talented and qualified teams and technicians that can go and address end customers.

He also wants Treeline’s to focus on product surface area so that it can more easily service more customers and customers from different industries than a traditional MSP that would need more time for skilling up, hiring or buying another MSP business. He wants no Treeline customer to outgrow its capabilities.

“We want to be the most forward-thinking IT service provider,” he said. “There will be no reason for anyone to ever work with a different type of MSP versus the one that we’re building. And that will further compound and allow us to grow faster and have a higher customer NPS (net promoter score) and also flow into the bottom line, because we’re doing it more efficiently with more products than anyone else is offering.”

Joining Doyle at Treeline is co-founder and Chief Technology Officer Hussain Kader. The CTO previously worked at Brex for about three years, leaving the finance platform provider with the title of senior software engineer. His resume includes about five years with Menlo Security, leaving in 2020 with the same title, according to his LinkedIn account.

Another notable member of Treeline’s C-suite is Jeff Gaines, chief growth officer. Gaines previously served as senior vice president of growth at Lyra Technology Group and was CEO of solution provider Interlaced, according to his LinkedIn account.

Here’s more of what Doyle told CRN about the new kind of MSP he and his team are building. Answers have been edited for length and clarity.

What led to you founding Treeline?

My background was in engineering. (That’s) what I studied in college. It’s where I met my now co-founder and Treeline CTO Hussein (Kader), who’s a much better engineer than I was.

He went the software engineering route and worked at really strong companies (including Menlo Security and Brex).

I went the venture capital route. And I joined a venture capital firm (Accel) right out of school. And everything that I was investing in while there was IT infrastructure software, cybersecurity software, a lot that would just naturally sell to the enterprise.

I kept getting more and more interested in how SMB (small and midsize business), midmarket–lower-middle, upper-middle market companies–manage their IT and security, buy and leverage different pieces of modern software or legacy software and how it all works.

They often don’t have the resources or internal knowhow to manage it all. (Very) quickly, you run into the managed service provider space, which has been, at least from the Silicon Valley perspective, hiding in plain sight for a long time.

Modern tech companies didn’t have as much familiarity with the MSP model despite the vast majority of U.S.businesses working with MSPs.

The core insight that we had when we started Treeline was that MSPs modernizing in any which way they can–remote work, cloud, moving into cybersecurity, whatever the next wave is–it’s really hard to when you’re low-resourced or when your only approach or solution is to buy more third-party software and SaaS (software-as-a-service).

Our view, though, is that the pace of technology today–and now AI, general automation software–is moving so quickly that we just don’t think that the industry as a whole is necessarily prepared for that or able to keep up with it through the purchasing of more and more software and modernizing their existing software.

We’re setting out to try to build what we perceive to be the modern or next generation service provider model where we still have technicians, expertise, humans in the loop. There’s so much longtail and scope that’s required to service customers well.

But when and where can software automation and AI very securely and carefully augment that technician workflow so that we can better service customers in a faster, more secure, cost-effective way, but also just scale our business better so we’re not hitting natural constraints that I feel like the industry is always hitting as they try to scale and service ever growing customers.

How’d you get connected with Andreessen Horowitz?

Venture (capital) and Silicon Valley as a whole has gotten a bit more scrutinizing on the business models they’ve been investing into for a long time.

You see a lot of the common points of conversation online–is SaaS dead? What is AI going to become?

There are a lot of kernels of truth in there where we should be asking the question–what is the right business model for modernizing industries (and) serving end customers? Maybe it’s not yet another credit card swiping SaaS tool. Maybe it is a different model.

Our view is that this vertical integration–let’s become what we think that that entire next-generation service provider can be–is the right model for this industry. And you can debate it across so many others, we’ll see different business models come up for this industry. And Andreesen agreed.

I remember going to the partner meeting. And you have a lot of very experienced operators from older school places like VMware and different places.

They were just like, ‘We’re investing in too much software in this day and age.’

I’d known (Andreesen partner) Joe Schmidt–who is on our board now–for quite some time through the venture world. But just talking to them and being in that room where folks are not asking rudimentary questions on, ‘Why are you not doing the business model this way? Why can’t you just do it like these other folks?’ It was more an encouragement that we actually could build a cool company this way.

How much of the MSP business have you been able to fully automate?

We entered into it being like, ‘Let’s automate away everything.’ Not disintermediate entirely technicians, but they’ll focus on other higher-value tasks, which still can be true.

But having people, technicians support relationship building in the loop with everything we’re doing is essential. There are some workflows and agents we can build and different features where we can drop this pesky 10-minute ticket down to zero. No one needs to look at it. That’s great. But it’d be foolish of me to say that we can do that across the board.

These are complex environments, and any 1 percent likelihood error can be a big issue. So we shouldn’t actually have the North Star goal of trying to zero out everything.

What’s more important is (this) 50-minute task that required a lot of escalations back and forth with the customer–why can’t it take five minutes? Where the technician is doing that core four-minute task and (then using) excellent documentation, context augmentation.

We’ve already spoken to the customer, clarified it or read something from their agent and machine that gives the piece of information we needed. Like, let’s drop that 50 down to five so the technician isn’t (trying) to harvest all this information and manage all these different levels of technicians and the customer.

That is the core value prop of our model. Because I don’t want customers to just be like, ‘Oh, now we’re talking to AI.’ I don’t think things are there yet. Nor will they be there for a long time. They should still have that white glove support. But behind the scenes, we can just clean up a lot of really complex processes (and then) free up technicians for more interesting, higher value work.

So will Treeline build traditional MSP tools?

We are not going to build our own MDM (mobile device management) or EDR (endpoint detection and response) solution. CrowdStrike and SentinelOne, and all these companies have armies of people.

We do need to partner with these folks, integrate with their products, collect data from them. There will be some pieces of the, let’s call it, more legacy, existing tooling stack that we can augment or replace and build ourselves. But it’s less, can we just knock out a bunch of tools that people use and put it into our own thing? It’s, what is that really value-added thing that the industry just hasn’t seen yet versus let’s just build another one of these because we think we can do it well.

While a lot of our focus and discourse is around AI and what it can do for this industry, AI is one input. (And) it’s a powerful input. For our first nine months in the industry, we used AI to write our software.

But we just built good infrastructure, data services, plumbing. We could have gone in, day one, and scripted a bunch of things and put out some agents. But that’s not scalable. I don’t think that’s very differentiating.

For the first almost year of the company, we’re integrating with these legacy tools. We’re building infrastructure so that we can then leverage AI agents and just good software to fundamentally change this industry. So it is going to be fast paced.

We are building something very unique very quickly. We do have to ground ourselves in how this industry operates and the tools that they’re used to using.

Is Treeline displacing people already?

This industry has enough people displacement folks out there. The traditional rollups and this and that.

You do hear some horror stories around anytime you hit scale (that) there’s concern around disintermediation of technicians or whatever role.

We are fully grounded in growth. Yes, we want to implement really good technology and really augment that technician in really cool ways. But if we aren’t focused on growth, it’s a tougher model.

We can scale this very well. And that further empowers the technician teams we work with. But if we didn’t have a growth lens, or just we’re looking at margin expansion, I don’t think we would be much different. It’s just a different tool for the same stuff that you see deteriorating the industry.

This conversation, of course, is, like, very focused around the technology we’re building and all of that, and that’s a huge emphasis. But if we’re building something differentiated and we’re freeing up experienced technicians’ time for stuff that they should be doing versus the more distracting stuff, let’s talk to the end market, convince them that our solution is the right modern one for them and really grow.

How big is Treeline today?

We have a couple hundred customers and an employee team of about 70 to 75 folks.

That includes technicians in the loop, experienced MSP operators, the go-to-market team we’re building. But then also the engineering team that we’ve built and that Hussain is leading.

These are world-class engineers addressing a problem in an industry that I don’t think has been the attention of engineers that have that background. I’m really excited about the strength of engineering that we’re bringing into the fold.

We’re growing very well and shooting for 3x-plus revenue growth this year and trying to scale rather quickly. I’m able to confidently say these lofty growth numbers just because we’re hearing from the market that what we’re putting in front of folks is really resonating.

I’ve seen the cautionary tales of growing at all costs and earning a ton of money and doing all of these things that Silicon Valley honors and glorifies. Yes, let’s grow quickly. And let’s try to really reinvent this category. But do it responsibly, ideally profitably, in a careful way where–along this value chain that we are trying to disrupt in many ways.

Not disintermediate any persona, if you will. Just the legacy processes and some tools and assumptions that the rest of the world makes around what the MSP space actually is.

How are you investing this money?

Most of the investment is engineering and product. (We’ll) get deeper and deeper into the problem space and have so many different areas we want to build.

We also are very meaningfully expanding and building our go-to-market team. We just recruited Jeff Gaines (as chief growth officer). He’s a long-time industry veteran (who most recently served as senior vice president of growth at Lyra Technology Group).

He joined us in January and has built in short order here an exceptional go-to-market team, philosophy, stack.

We are very focused on organic growth, and Jeff is leading that charge and excited to continue to make those investments in both go-to-market and engineering. (And) we are profitable today.

What does your customer base look like today and where’s it going?

When we incorporated the company, essentially, we pretty quickly merged with a couple of traditional service providers. Ones that we felt had extremely strong technician basis, good operational expertise.

I knew the category from another vantage point. But I’m not a 20-year industry operator. And so, we need to build that expertise and that muscle into the company while also learning from the right people, from the right technicians, from their end customers.

While we have some slight inorganic growth motion through finding these right service providers to partner with, over time–and starting already now–we’re leaning more so on organic growth. And we’ll continue to opportunistically look for these folks and these businesses that are special, in our view, for criteria that’s different from the typical private equity company out there.

When and where can we inject Treeline with the industry’s best and brightest on the technician and operating side? We are not going to SaaS-ify what we’re doing. We need people in the loop. And we need to ground ourselves in how the industry operates versus bringing in hubris that, perhaps, other companies could be (bringing) saying, ‘Oh, we know what to do.’

(Our customer base has) been guided a bit by those initial technician teams and businesses we brought on. That was a good constraint because we don’t want to be everything for everyone.

But through that motion and through how we built our offering, we do have a good customer base across CPG (consumer packaged goods), manufacturing, health care, some financial services.

The modern venture-backed tech business world as well. Geographically, we have most of our offices and our folks on the West Coast. We certainly have an East Coast presence as well. But, we find that our customer base naturally started a little bit more West Coast.

We’re able to service the rest of the U.S. very well just by nature of where the industry is today. Folks don’t necessarily need us driving to them every three days. But we have enough folks distributed to service the large metro areas.

What’s Treeline’s business model?

We’re watching the discourse around the metered outcome-based (approach).

It is a function of meeting the industry where they are. Just trying to price it in an understood way around per user while also offering different projects and ancillary offerings around the core recurring piece.

For now, fairly standard pricing, hopefully more competitive pricing just because everything we’re doing on the back end should certainly improve service.

We’re leveraging software versus continuing to add a ton of new people. And so, there’s a cost element, too, that we’re really trying to lean into as well.

Is it easier for Treeline to show customers the impact of your products and services compared to traditional MSPs?

That’s a nice byproduct of us having started by building data services and infrastructure and building everything ourselves. It’s not just exposing it to technicians.

Client success–what do you need to speak to the customer in a way that’s surpassed the tried and true, yet manual, slow cadence QBR (quarterly business review).

You as the customer, your interface for this industry is typically maybe some forms, but you’re picking up the phone. What interfaces can we build to the customer that, again, doesn’t disintermediate the human relationship, but gives them the information that they want to see in real time?

That’s the huge benefit of our model. And it’s impossible for us to sell that to the industry because that scope is huge. And we’ve built around infrastructure. But we can expose it to all these different personas, both customer and operator.

What technology vendors do you consider close partners?

We try to be careful with sprawl just because we need to integrate pretty meaningfully with the ones that we do integrate with.

Microsoft and Google, depending on the type of customer. We’re building partnerships with CrowdStrike for some and SentinelOne for others.

We do want to be prescriptive with what tools folks are using, but not overly so where if they’re a SentinelOne shop or whatever it might be it’s a discussion versus something that needs to be changed.

So (we are) Microsoft heavy. We built out some further integrations into what I call the modern technology stack–like Google, Okta, Apple device managers–versus just (Microsoft’s unified endpoint management service) Intune.

The core back end of what we’re building is able to like, be extensible enough to not say, well, we’ve just only built our piping to this unique Microsoft type of license. It is more extensible than that.

How are you presenting Treeline’s advantages to customers?

The framing that we have is, yes, we are trying to innovate meaningfully in this category, to leverage modern tools and really affect a lot of change in the space, but always in pursuit of pulling the market forward with us.

If we’re to speak to end customers very effectively and tell a story that is very unique, but it’s reassuring to them that these modern technologies are only a benefit versus a risk, we’re pulling them forward with us, meeting the industry where it is.

Accelerating its modernization is the continued pursuit we have. That’s why we started basically as a services business, through bringing those technicians and teams on and over time, as we continue to build excellent software and inject AI carefully into automation software, we’re pulling it forward versus coming from a completely different lens, saying, here’s something brand new, maybe it’s like shiny, but it’s something that is disruptive to the industry.

As powerful as AI and modern technology is, it’s just as hard to implement and use well. And it should come with associated concern around–is it secure? Can we use it effectively?

And we’re trying to solve all those problems. The way that we use AI and any newish type of software application in this space needs to be wrapped extremely well in security and be near-risk free with how we’re using it.