Custom Automation Platforms, Vertical AI Exits & What I'm Reading
Capital Efficient #3
I’m Pat McGovern and welcome to the third edition of Capital Efficient. Let’s get into it.
Weekly Radar
The Rise of Custom Automation Platforms
The last few weeks have seen the public rise of a new category of software companies that don’t really have an agreed upon name yet - for the sake of simplicity, let’s call them Custom Automation Platforms. Every Fortune 500 company is strategizing around how they can implement AI in their business to save money, drive topline growth, and maintain their competitive edge.
But for the largest enterprises, most vertical AI startups are not broad enough or malleable enough to slot into their idiosyncratic workflows. And vertical AI platforms can struggle when pitted against decades of calcified internal processes and procedures.
So how do large companies solve this? Historically, when these businesses can’t find an off-the-shelf piece of software to automate something or surface some business insight, their alternative has been to hire consultants to build internal tools. Evidence of this can be seen in the oft-cited reports around Accenture generating billions in “AI” bookings. However, the products that these consulting engagements produce are rarely best in class.
Now, a third path is emerging: the rise of custom automation platforms (CAPs). These CAP startups look to automate some of the most complicated workflows in the largest and most demanding enterprises (think top healthcare payors, CPG brands, hardware manufacturers, etc.)
These businesses have a heavy forward-deployed aspect and start off looking like consulting , with the hope that over time they can build a core technical asset with enough reusable automation components to allow them to scale like a high-growth software startup. With the cost of code generation falling exponentially, it makes this CAP model, which would previously have been too costly and time intensive, look instead like an exciting new way to deploy AI in the enterprise. CAPs are fundamentally new types of software businesses that have only now been made possible by advances in AI.
And custom automation platforms are having a moment. There were three big developments in the CAP space over the last few weeks:
Distyl: Distyl, the CAP that has been in market the longest, announced a $175MM fundraise which valued the business at almost $2B - a 9x jump from their Series A valuation. This round was led by top funds like Lightspeed, Khosla, DST Global, Coatue, and more. Some examples of Distyl’s recent custom automations include an AI system for PA decisioning built for a top healthcare payor and an AI insights engine for mapping demand and supply chain changes facing a top OEM.
Brain Co.: Brain Co. came out of stealth with $30MM in funding led by Elad Gil and Jared Kushner, as well as a bevy of tier-one angels. Brain Co. describes themselves as building “AI for the world’s most important institutions.” They are intentionally trying to stay broad and avoid verticalization, instead looking to serve a wide variety of industries that share common challenges. Like Distyl, they are taking an early view towards healthcare and supply chain as places where they can make early inroads.
Percepta: Not to be left out, shortly after the news of Brain Co.’s raise broke, General Catalyst announced their own newly incubated “AI transformation company” called Percepta which will focus on “transforming critical institutions using applied AI.” Percepta is swimming squarely in the same lane as Brain Co. and Distyl. They are hiring Applied AI Engineers in the US and Europe and are targeting large organizations in the “industries that power the world” including energy, healthcare, and manufacturing.
Using Consulting to Build Tech
All of these CAPs are doing unscalable things for the moment in hopes of building out their core technical platform, which should improve as they successfully automate more and more enterprise work. For now, much of what they do doesn’t look different than consulting, but each of these projects creates more reusable components for their proprietary platforms.
In terms of their flagship technical asset, Distyl’s technical platform is called Distillery, while Percepta offers a core asset called Mosaic, which enables enterprises to “deploy production-grade, industry-specific AI workflows in minutes instead of months.” All of these seem to draw inspiration from Palantir and their Gotham/Metropolis platforms.
One big question in the CAP category is how many of these workflows get built by forward-deployed engineers vs. customer employees, and who is responsible for maintaining them in the long run.
The CAP space will likely produce mulitple large winners, and while the very top of the market (e.g., F200) is getting a little crowded in terms of VC-backed CAP startups, I think there are still many opportunities in this space. If you are building a Custom Automation Platform, I’d love to hear from you.
News Round Up
Prepared, a public safety vertical software platform that had raised $150MM from funders like a16z and General Catalyst, announced it would sell to Axon, a leading public safety vendor best known for its tasers and body cameras. Axon has already been selling municipalities a GPT-wrapper called Draft One for writing police reports and this partnership should be a big unlock for Prepared given Axon’s existing relationships with nearly every hypothetical Prepared customer.
This is seemingly an early exit for Prepared with the business having raised an $80MM Series C in Q2 2025. My speculation is that selling AI to the government is still fairly painful, and they had to cut a deal with Axon to get access to customers at scale. This makes me wonder about the fate of many other public safety AI companies that have raised in the last few years using Prepared as a rapidly growing comp. Prepared was reported by some outlets as having sold for ~$800MM but there’s been no official confirmation of that number. For context, Axon’s market cap today is $56B so that scale acquisition would be a mere drop in the bucket for them. Axon generated $670MM in Q2 2025 revenue, $290MM of which came from their rapidly growing software business.
EvenUp @ $2B. This summer, Harvey - the leading legal AI platform for large law firms - announced a Series E at a $5B valuation with reports surfacing that the business had surpassed $100MM ARR. Now the plaintiff’s bar is getting in on the act with EvenUp announcing a $150MM Series E at a $2B valuation. EvenUp targets plaintiff-side personal attorney firms and has been expanding its offering after establishing a foothold drafting automated demand letters. An impressive valuation given the size of the law firms it targets.
What I Am Reading
Ethan Ding’s Substack (a/k/a Mandates) - I really enjoyed two pieces in Ethan’s Substack - the first on OpenAI’s potential ads strategy and the second on the agent builder / cloud provider tension. Both are must-reads and very thoughtful.
ICONIQ State of Software 2025 - ICONIQ has released their latest State of Software report - always worth a read to get a pulse on the market.
Really solid read. I am building a salesforce native Custom Automation Platform. Let me know if it's worth a chat!