When I first met Ankur Patel over a year ago, he had already crossed $1 million in ARR with zero equity fundraising. As he put it, he was “bootstrapping like it was nobody else’s business.”
It wasn’t his first rodeo. His last company raised more than $50 million, and from that he learned the downsides of raising too much money too quickly.
“AI is super hot. Funding is easy. And as a founder, it’s tempting to think raising capital is the right thing to do. But if you’re not careful, it leads to unforced errors.”
Today, Ankur leads Multimodal, one of the most successful companies in Generative AI automation for insurance and finance. We talked about the shift from RPA to GenAI, the real limits of AI today, and how highly regulated industries can adopt these technologies without losing trust.
Ankur is clear-eyed about what makes this moment different:
“RPA are software bots meticulously designed to perform certain actions precisely the way they were programmed. The challenge is most work is variable — you need human-level reasoning to automate it. RPA just wasn’t suited for that.”
With large language models, that variability is finally addressable.
“With the latest generation of GenAI and Agentic AI, models are very good at understanding language. They can handle variability really well. That opens up use cases that were previously not automatable but now could be.”
This isn’t about cutting corners. It’s about creating entirely new workflows where decisions that once took weeks can now happen in minutes.
Selling into finance and insurance means technology alone isn’t enough. Trust is everything.
“We don’t view ourselves as just a technology company. We’re in the business of trust.”
That trust is built step by step:
“It’s easy to get to 90%. Maybe even 95%. But in finance, a 5% error rate is catastrophic. That’s why our forward-deployed engineering matters so much — we close that last mile.”
And importantly, the goal isn’t headcount reduction.
“Most of our customers aren’t looking at this as a way to cut labor. They want to make their people better off at serving the end customer.”
In head-to-head evaluations, Ankur sees the same themes emerge:
“Enterprises spend significant time onboarding new vendors. They want to be sure the partnership is off to a flying start. What wins is forward-deployed engineering to deliver outcomes, and the ability to absorb expert feedback to drive high-precision workflows.”
Rather than spreading thin across many low-value tasks, Multimodal leans into the hardest, most business-critical use cases.
“We’re not trying to be a general-purpose automation platform. We focus on high-precision workflows where tolerance for error is zero — and we aim to be the best in the world at that.”
While many AI startups race to raise bigger rounds, Ankur is more deliberate.
“From the very beginning, we knew we were selling something new into a space that doesn’t buy quickly. We wanted a really good foundation before bringing in external capital to scale.”
Bootstrapping past $1M ARR wasn’t an accident — it was by design. Now, capital will be raised only to serve repeatable demand, not to manufacture growth.
Looking ahead, Ankur is confident about where the market is going:
“What you’re going to see is implosion of point solutions. Enterprises don’t want to manage dozens of vendors and integrations. They want platforms. AI is going after the labor budget, not just the software budget. That makes the opportunity an order of magnitude larger.”
In his words, Multimodal is building the “system of intelligence” that sits on top of existing systems of record, quietly doing the heavy lifting while letting organizations keep the processes they already know.
Ankur Patel’s journey reinforces a few timeless lessons:
I loved this episode of Under the Banyan Tree. If you want to understand how GenAI is actually transforming insurance and finance — and what it takes to adopt it responsibly — give it a listen.
Youtube
Spotify
.png)