We Raised $30 Million to Build the AI Operating System for Regulated Industries

When we founded Notch in 2021, we were not planning to build a software company.

We were building a specialty insurance product for digital assets. One of the first of its kind, operating under strict auditability, documentation, and policyholder communication requirements. We wanted to use AI to improve how we served customers and ran operations.

What we found was that none of the tools available could meet the bar we needed. Existing chatbots could deflect tickets but they could not safely complete end-to-end workflows with deterministic controls, hard limits, and a clean audit trail. Every AI tool we evaluated was a black box: unpredictable, hard to audit, and not built for an industry where a single mistake carries compliance, financial, and legal consequences.

So we built our own. An internal operating layer that combined conversational AI with structured execution logic, permissioning, validation, and mandatory escalation when policy or data was unclear. It worked. And we realized the broader market needed it.

Today, we are announcing that Notch has raised $30 million in a Series A funding round, bringing total funding to $45 million. The round was led by Headline, with participation from Lightspeed Venture Partners, Jibe Ventures, Illuminate Financial, and Phoenix Insurance.

‍

The problem we set out to solve

AI in regulated industries has a production problem.

Pilots work. Demos look great. But when enterprises try to take AI from proof of concept to day-to-day operations, they hit a wall. Compliance teams want auditability. Legal wants hard limits on what the AI can commit to. Operations want consistency across regions and channels. And all of it has to sit cleanly on top of legacy systems that were not built for AI.

The result is that most "AI deployments" in regulated industries are either surface-level chatbots that deflect tickets rather than resolve them, or heavily scoped tools that handle one narrow workflow while everything adjacent still runs manually.

We built Notch to solve a different problem. Not a chatbot. Not a task-specific tool. An AI operating system: a platform that lets regulated businesses deploy agents across conversational and back-office workflows, with the governance, auditability, and control their environment requires.

‍

What Notch does

Notch deploys AI agents that resolve issues end-to-end.

On the conversational side, that means broker and policyholder interactions, policy servicing, structured intake for claims and underwriting submissions, and co-pilot functionality that lets adjusters and underwriters query complex documents in natural language and receive structured, traceable answers grounded in source data.

On the back-office side, Notch automates high-volume operational workflows: ingesting documents and communications, extracting structured data, classifying and routing claims, and surfacing time-sensitive requests before deadlines are missed.

The controls that make this safe in production are not a single feature. They are a layered architecture: conversation safety checks, defenses against adversarial misuse, hard access rules tied to identity and verification, business limits with deterministic thresholds, and jurisdiction-aware rules that adapt to local regulatory requirements. Every action is auditable. Every escalation is logged. The system does not guess in high-stakes situations.

‍

What the past year taught us

Over the last 12 months, our ARR grew 12x. We deployed with leading global insurers and financial services firms. We built a body of evidence about what production AI in regulated environments actually requires.

The biggest lesson: the gap between a working demo and a production deployment is almost never about the model. It is about everything around the model - the permissioning, the escalation logic, the audit trail, the jurisdiction handling. Getting that right is an engineering and design problem as much as an AI problem. It is where most vendors stall. It is where we have spent most of our energy.

‍

Why now

We were deliberate about who we raised from.

Headline leads this round as our primary institutional partner. Lightspeed Venture Partners, Jibe Ventures, and Illuminate Financial each bring deep experience in enterprise software and financial services. Phoenix Insurance joins as a strategic investor with direct operational insight into what regulated insurance workflows actually require at scale.

Gil Tamir, Phoenix's Deputy CEO and Director of Innovation and Technology, put it clearly: "The hard part is building systems that meet governance and reliability requirements while also improving the customer experience. You need to be able to explain every decision it makes, keep it under control, and trust that it is consistently accurate. Notch has created a system that does just that."

That is the bar we hold ourselves to.

‍

What's next

This funding accelerates two things.

First, our expansion in the U.S. The U.S. insurance and financial services market is large, complex, and at an early inflection point on AI adoption. Regulated companies are moving from evaluation to deployment, and the infrastructure they need to do that safely is not yet widely available. We intend to be the platform that closes that gap.

Second, continued development of the Notch platform. Deeper workflow coverage, more sophisticated orchestration across multi-agent pipelines, and the enterprise infrastructure that large carriers and FSI firms need for long-term commitments.

We are also hiring. If you are building production AI systems for enterprise environments - on the product, engineering, or go-to-market side - we would like to talk.

‍

Thank you

To our customers: you trusted us with real operations in high-stakes environments. You gave us the feedback that shaped the product and the evidence that the approach works. This milestone belongs to you as much as it does to us.

To the Notch team: building production AI for regulated industries is one of the hardest problems in enterprise software. You did not cut corners on the parts that matter. That is rare. I am proud of what we have built together.

To our investors: thank you for backing a long-term thesis in a market that rewards patience and precision. We intend to deserve that trust.

The foundation we have built over the past five years - built inside insurance, stress-tested in production - is what makes us confident we can execute on what comes next.

‍

Rafael Broshi

Co-Founder and CEO, Notch

‍