Jason Boehmig, founder and CEO of Ironclad, breaks down taking risks, disrupting categories, and building big value for customers with AI
Jason Boehmig, founder and CEO of AI-powered contract management software company Ironclad, discusses his top learnings and insights from building a successful AI-first company from the ground up with AssemblyAI’s founder and CEO Dylan Fox.
Founded in 2015 by Jason Boehmig and Cai GoGwuilt, Ironclad is a contract lifecycle management (CLM) platform trusted by thousands of legal teams to innovate their digital contracting.
As an attorney by trade, Jason realized that there was a gap in the legal market for modern technology to assist with filling out detailed, technical documents and ensuring that the details accurately populated to other necessary documents—so he started playing around with building his own workflow automation. Initially, Jason’s goal was simply to become a more efficient attorney, but that quickly morphed into a much larger vision.
After quitting his job as an attorney to focus on building automation software full time, Jason connected with his co-founder, Cai, who had also recently quit his job as a software engineer. Together, Jason and Cai joined forces to build Ironclad into the $100M ARR company it is today with customers like L’Oreal, OpenAI, Cisco, Shell, Invoca, Plaid, and Quora.
Jason recently sat down with AssemblyAI’s CEO and founder Dylan Fox to discuss how AI was a core part of his original vision for Ironclad and how that vision has changed and matured over time, as well as when risk-taking should be encouraged or tabled, how to create big value for customers, and much more.
Here are a few takeaways from that conversation:
1. A winning product strategy solves for end-to-end workflows
Jason Boehmig: “The idea around Ironclad from day one and still continuing to this day is those two parts of the contract: making the contract and then managing the contract after it's been made. What we realized was workflows were a better way of making contracts than just an AI that's on an email alias. Because even if that AI would get something wrong 5% of the time, you really can't have that in a business contract. Like, whoops, we put the wrong information on page 37, and you didn't catch it, sorry about that. It doesn't really work. So we had to be 100% accurate. And the way we could get to 100% accurate on the contract creation was through really great workflows.”
“And of course, once you have a workflow, at the end of that workflow, you have a bunch of really great structured data. Incidentally, adding workflows to this old school software category called contract lifecycle management was what helped us, in the business school sense of the word, disrupt the category.”
2. The biggest breakthroughs come by focusing on the customer, not the product
Jason Boehmig: “We started saying, okay, there's a potential breakthrough here. It's not necessarily usable yet, but we need to start seriously thinking about the Ironclad.ai piece here, because we're making contracts great in our workflows, but if we can get the ability to extract from 10,000 PDFs all of the data and put them in our database as well—that's going to generate a ton of value for our customers. And so we started retooling with the idea of layering in AI specifically at that point into just the data extraction.”
3. Companies should take on the risks, not their customers
Jason Boehmig: “I think I would distinguish between risk tolerance that we have as a company and risk tolerance that we have to put an experience in front of our customers. I'm comfortable taking a lot of risk, and I think that's one of the benefits of a founder-led company, but I don't want my customers to be taking any risk in using the product.”
“There's a couple of ways that shows up. One, we're rigorous in testing. So there's been a lot of noise around AI products and standalone AI products in legal space, and we have really focused on our flagship product in terms of the public narrative. We are in rigorous testing and have been for some time around our standalone product…but you could say we’ve been slow to launch that”
“We have a really high bar for what we're going to put our name behind in the industry.”
4. Have a strategic point of view on when to build from scratch—and use state-of-the-art AI models for everything else
Jason Boehmig: “Our current view, which seems to roughly match with the large model providers, is that the verticalized application stuff is things that we're going to need to be uniquely good at and we're going to develop our own proprietary stuff around.”
“Whereas the foundational capabilities, we should stay on the latest state-of-the-art and invest in anything that allows us to stay on the latest state-of-the-art. I'd say specifically with respect to the foundational model companies, our approach is to constantly be evaluating that.”
5. The AI strategy that sounds compelling isn’t always the right strategy to pursue
Dylan Fox: “How has your technology strategy—which is like just stay on the main branch, be creative, focus on the applications and the user value—fared out, would you say, compared to the competitors?”
Jason Boehmig: “I would say like from a strictly technological standpoint…extraordinarily satisfied…11/10 on that one.”
At first…” I was like, hmm, that sounds pretty good—like to be on your own fully custom-trained model,...And I do think there's a marketing message which is we have an LLM that's completely custom trained for you, that does land…so I think the market is still playing catch-up to the reality of the results.”
Dylan Fox: “I think what's tough for earlier stage founders too is that that story also sometimes can sound more compelling to investors. Like, hey, we're gonna go create a specific model for X. But to your point, that might have an advantage for only 3 to 9 months and then you’re worse off...”
6. Don’t lose sight of the big picture in your quest to innovate
Jason Boehmig: “I do think it's really important to have a multi-product strategy if you're an incumbent and you can't lose the direction of your main product— there's so much momentum behind that. If you're an incumbent, it's an existing category. You should have AI everywhere, but don't lose sight of the big picture on your main product.”
7. The state of AI innovation is constant, but it’s also energizing
Jason Boehmig: “Thinking back to nine months ago feels like almost as long ago as starting the company, which was ten years ago, like a totally different state of the world. And it's also giving me a ton of energy as a founder. And as I talked to other founders who are kind of like ten years in, it's a grind, you know, to be a founder. But having this boost of energy around absorbing new information every day, trading notes with other founders, getting hands-on with a new product. That's been really fun.”