Today we are announcing that ancep has raised $8.5m. Creandum led the round, with Kleiner Perkins and Vicus Ventures participating, alongside angel investors including the former CFO of Aon, the former Chairman of Marsh, the former head of insurance of Palantir, as well as operators from OpenAI, Anthropic and Google DeepMind.
We build AI agents for commercial insurance brokers. They help brokers place the risks they currently turn down, and earn revenue they would otherwise never book.
The risks brokers turn away
Placing a risk takes roughly the same effort whether the commission is a few hundred pounds or tens of thousands. So below a certain size, brokers decline the work or leave it at the bottom of the queue. We estimate that up to 30% of the risks a broker sees fall below that line, and that about half of them could be placed profitably. That is 5 to 10% in revenue a large brokerage never books, which our agents help recover within a few weeks and without adding headcount.
The timing matters. Global commercial insurance rates fell 6% in the second quarter of 2026, the eighth quarterly decline in a row, as capacity stays abundant and capital in surplus.
Plenty of vendors have offered brokers time saved and cost reduced. Those pitches tend not to resonate, and they get stuck in two-year procurement cycles. What we sell is net new revenue: business a brokerage sees today and cannot afford to place.
What our agents do
Our agents carry out the routine stages of a placement. They read the client’s submission, build the broker’s standard documents from it, and approach insurers for terms by email, portal or API.
They work inside the platforms brokers already run, including policy admin systems, e-trade portals and CRMs, so there is no integration to build for each one. The largest brokers often run dozens of policy systems side by side. For them, that is the difference between a rollout measured in days and one measured in years.
We deploy one workflow at a time, starting where the pain is worst and the benefits come quickest. We show a working prototype within two days of signing an NDA. We charge a portion of the value our agents create, so we are paid only when the broker is.
The broker stays in charge
Only an authorised broker can bind cover, and the broking team reviews every agent output before it takes effect. Each client runs in its own isolated environment, and nothing is pooled or trained on across brokerages without explicit consent.
Before ancep, our founders sold AI to the testing, inspection and certification industry, where an error can put physical infrastructure at risk. We build ancep to the same standard.
Already live in the US, UK and Europe
Our agents are live with A.C.I.C. Financial Advisory, one of the largest credit insurance brokers in Austria, with a top 10 UK broker, and with the specialty lines of one of the top US brokers. The chief operating officer of one customer division reports a 10% revenue increase within six weeks of going live.
A.C.I.C. specialises in credit insurance and surety. Its brokers used to spend much of their week reading complex policy wordings, comparing insurer terms line by line, preparing market submissions and pulling information out of large document sets. Our agents now support that groundwork, and jobs such as detailed contract and policy review take a fraction of the time they did. The firm has taken on more clients without adding administrative headcount.
“We did not do this to cut costs but to get our specialists’ time back.”
Christoph Zawadil, Managing Director at A.C.I.C. Financial Advisory: “Credit insurance and surety is specialist work, but a lot of a broker’s week goes on comparing insurer wordings, preparing market submissions, extracting information from large volumes of documents. ancep understood that workflow deeply, rather than trying to sell us generic AI, which is why adoption across the business has been exceptionally high. We did not do this to cut costs but to get our specialists’ time back, and they are now using it to advise clients, expand into new markets and build insurance products we didn’t have the capacity to design previously.”
Building for brokers
Plenty of people are betting that brokers get disintermediated. We think brokers have huge leverage, and no chatbot is going to replace it. So we are building for them, and we are starting with the twenty largest in the world.
Our investors see it the same way.
Simon Schmincke, Partner at Creandum: “Commercial insurance is a $3 trillion market that has absorbed ten years of software promises without much changing inside the largest brokers. Gagan and Jake built something that works within the systems those firms already run, meaning they are live and expanding across lines of business within weeks. Most AI in this market is a bet against the intermediary, but ancep is the first company we have seen that makes it more valuable instead. We’re excited to back them in that mission.”
Mamoon Hamid, Partner at Kleiner Perkins: “Commercial insurance is a large market where too much business still goes unserved because the economics don’t work. ancep is using AI to change that equation while keeping brokers at the center. We’re excited to support Gagan, Jake, and the ancep team as they expand what brokers can profitably serve.”
Where we come from
Our founders, Gagan Khurana (CEO) and Jake Turner (CTO), met at Oxford. Previously, they spent a year together at Scope AI, an AI provider to the testing, inspection and certification industry. Gagan ran sales and Jake ran engineering as revenue scaled rapidly in under a year and the company raised a $20m Series A from Index Ventures.
Gagan arrived in Britain from Kabul as a refugee in 2001 and sold on his father’s stall in Shepherd’s Bush Market before joining McKinsey, where insurance engagements first put him in front of brokers. Jake was the first engineer at the quantum computing company QuantrolOx, and interviewed more than 50 brokers in depth during an earlier venture.
What comes next
We will use the new funds to support growth and hiring. Thank you to the brokers who put our agents to work on real placements, and to the investors backing us to do more of it.
