Our investment in Notch: The Operating System for Regulated Customer Ops
We are thrilled to announce our participation in Notch’s $30m Series A, led by Headline, with participation from Lightspeed Venture Partners and Jibe Ventures.
At Illuminate Financial, we invest in the unglamorous but essential plumbing of the financial system. We look for category-defining companies transforming infrastructure where accuracy isn’t a “nice-to-have” — it’s the only option.
The “GPT Moment” for enterprise software has arrived, but it has hit a wall in regulated industries. While retail brands use LLMs to suggest summer dresses, financial institutions are stuck. For a bank or insurer, a “hallucination” isn’t a minor glitch; it’s a regulatory breach.
That’s why we’re excited to partner with Notch. Notch isn’t just another chatbot; they are building the Operating System for Customer Operations.
The Team: High-Stakes DNA
We back teams first, and the Notch team is the “Alpha”. The founders — Rafael, Yuval, Elool — have been operating together for over a decade, beginning in the Israeli Air Force (IAF). In a world where “close enough” isn’t an option, they spent ten years building software and managing operations where mission-critical accuracy was the baseline.
This team possesses a rare level of drive and velocity. They initially launched as an insurtech, but they had the conviction to “burn the ships” when they realized the true bottleneck wasn’t the insurance product itself — it was the operational layer of customer service across all regulated industries. They rebuilt from zero, found immediate product-market fit, and have since demonstrated explosive growth and impressive efficiency.
Our Thesis: The Shift to the “Context Layer”
As AI reduces the marginal cost of generating code toward zero, the value in software is shifting. We believe the next decade of enterprise value won’t be captured by the models themselves, but by what we call the Context Layer. Notch is building the platform that captures institutional knowledge — workflows, processes, and permissions — and turns them into executable intelligence.
1. Deterministic Control in a Probabilistic World
LLMs are probabilistic (they guess); regulated ops must be deterministic (they must follow the rule). Notch’s Policy Engine encodes a firm’s SOPs and compliance guardrails directly into the workflow. The AI operates within a “digital cage,” ensuring 100% auditability and control.
2. The End of “Human Middleware”
Financial services are plagued by “Human Middleware” — people whose entire job is moving data from an email to a CRM to a legacy mainframe. Notch moves beyond “chat” to autonomous execution. It doesn’t just talk to the customer; it triggers the back-office transaction, from processing a refund to settling a claim.
The $30 Ticket Wall
We focus on regulated verticals because the “pain” of customer operations is 10x higher than in retail.
- The Cost Gap: A retail FAQ ticket costs $2–$5 to resolve. In banking and insurance, complex tickets cost $15–$30, with high-touch disputes reaching $50+.
- The Compliance Tax: In insurance, CX already accounts for roughly 37% of total IT spending.
- The Productivity Unlock: AI has the potential to increase productivity in customer operations by 30–45% — but only if it can actually execute workflows.
By building specifically for these constraints, Notch creates a moat that “generic” AI cannot cross.
The Bottom Line
The AI CX landscape is a multi-hundred-billion-dollar market. While incumbents like Salesforce and Zendesk are adding “AI features,” we believe the winners in financial services will be purpose-built platforms that prioritize compliance and workflow execution over simple UI.
Financial services and insurance are the ultimate proving grounds for AI. If you can automate a complex dispute in a regulated environment, you can automate anything. We are proud to partner with the Notch team as they build the infrastructure for the next generation of autonomous customer operations.
