When my co-founder (and a very good friend) Deepak and I were working at Uber, we witnessed automation challenges at a scale that is hard to wrap one’s head around. I’m talking more than 200 million customer interactions in a single year.

We saw firsthand that while you need some degree of automation to handle that volume, the "human touch" is what actually solves complex problems. We left Uber with a nagging question: How can we scale support without sacrificing the quality and trust that only a human can provide?

Today, I am thrilled to share the answer to that question.

We’re proud to announce that Fini has raised $3.6M in seed funding led by Matrix Partners. We are also honored to have the backing of Y Combinator, along with some good friends from Uber, and high-profile angels from McKinsey and SoftBank.

This funding is more than a milestone for our company; it’s a validation of our core belief: 

AI support can (and should!) be held to higher standards than human support.

The "Trust Gap" in AI

We founded Fini in 2022 after graduating from Y Combinator because we identified a massive gap in the market. Everyone wants to use Generative AI for support, but in regulated industries like finance, insurance, and healthcare, the stakes are simply too high for errors.

We’ve all seen the headlines. From the Air Canada chatbot that invented a refund policy (leading to a lawsuit) to major financial firms like Commonwealth Bank of Australia and Klarna reversing AI support initiatives due to poor quality output. The industry calls these "hallucinations," but for a business, they are legal liabilities and PR disasters.

Deepak and I realized that until you can guarantee that an AI won't lie or violate protocol, regulated industries will always have to choose between efficiency and safety. When dealing with PII and ethics boards, safety almost always takes precedence. Which also explains why many healthcare and financial companies are leagues behind in AI adoption today.

We decided to put an end to this.

Elevating the Human Agent

At Fini, we built an agentic support system that monitors every AI conversation in real-time. But here is the differentiator: We built a Trust Layer.

Whenever our AI detects that its confidence score has dipped below a safety threshold, it doesn't guess. It alerts a human agent who reliably handles it. The AI learns from how that human handled the situation, ensuring that institutional knowledge is captured instantly.

This changes the game for the support workforce. As Deepak likes to say, "No human agent ever has to repeat themselves again."

We are moving away from the era where humans answer repetitive, cookie-cutter questions. Instead, we are empowering agents to become product experts who train and refine the system. We are turning the support function from a cost center into a profit center.

The Road Ahead

This investment from Matrix Partners, YC and others allows us to double down on this technology. We are already helping incredible companies like the US Chamber of Commerce Foundation, Covergenius, AtlasFin, and Bitdefender handle over 1,000,000 conversations a month safely.

I want to thank Patrick Malatack and the team at Matrix for believing in our vision to bring safe AI to regulated markets. I also want to thank our team all around the world, and our early customers who placed their trust in us with their most critical interactions.

We are just getting started on our mission to end AI hallucinations and make customer support faster, smarter, and safer than ever before.

We are launching Fini 2.0 and an AI voice support agent soon, but more on that later.

For now, from the bottom of my heart, thank you all for being a part of this journey.

- Hakim
on behalf of the entire Fini team 💙