The insurance-native AI that doesn't just answer the phone. It finishes the task - quoting policies, filing claims, updating endorsements, and hanging up.
There is a durable truth about insurance, which is that it is enormous, deeply regulated, and still runs, to a remarkable degree, on phone calls and hold music. Someone crashes a car, a pipe bursts, a policy needs a small edit, and the resolution to all of it is a human being with a headset navigating five internal systems while a customer waits. This is expensive and slow and, importantly, it is the kind of work that the technology industry has spent two decades politely declining to touch because it is neither glamorous nor easy.
Liberate, a San Francisco company founded in 2022, decided to touch it. Its bet is narrow and specific: build AI agents that live inside a carrier's actual systems and complete the whole task rather than answer a question and pass the customer to a person. The company likes to say it built "reasoning AI agents for insurance." What that means in practice is that its software will quote you a policy, take your claim from first notice of loss through the paperwork, and update an endorsement - and it will do this over voice, email, SMS, and chat, in English, Spanish, or French.
The framing matters because most "AI in insurance" turns out to be a smarter FAQ. Liberate's entire argument is that answering questions is the easy 20% and doing the work is the hard 80%, and that the value lives entirely in the hard part. In October 2025 a group of investors agreed with that argument to the tune of $50 million.
Figures reported by the company at its Series B, October 2025.
Liberate started with voice, which is a slightly counterintuitive choice. Voice is the hardest channel to automate - it is real-time, it is unforgiving, and a bad experience is memorable in a way a clunky web form is not. It is also the most valuable, because it is where insurance customers actually are when something has gone wrong. The company's reasoning was roughly: if you can do voice well, everything else is downhill.
Our agents don't just respond to queries or escalate tickets - they complete end-to-end tasks like quoting, claims, and endorsements.
The front door is a voice assistant named Nicole, which handles inbound and outbound calls, sells policies, and services requests. The company reports it resolves roughly 80% of the interactions it handles without a human, and it does identity verification and information gathering along the way. Behind Nicole sits what Liberate calls a "System of Action" - the agents that reach into rating engines, CRMs, and claims systems to make the change the customer actually asked for. A reporting dashboard hands the carrier call transcripts, sentiment analysis, and conversation-quality scores, which is the part compliance officers care about.
Inbound and outbound calls for sales and service, multilingual, with roughly 80% resolution and no human rep required for the common cases.
Agents that quote policies, process claims, and update endorsements by acting directly inside a carrier's core systems.
First notice of loss through the paperwork - identity checks, information gathering, and claim filing without a queue.
One layer unifying sales, servicing, and claims across voice, email, SMS, and digital channels.
Transcripts, sentiment analysis, and conversation-quality metrics so carriers can audit what the agents said and did.
SOC 2, HIPAA, PCI, and GDPR alignment - table stakes for selling software into regulated insurers.
The most telling thing about Liberate is where its founders came from. CEO Amrish Singh spent nearly four years at Metromile - the pay-per-mile insurer later bought by Lemonade - running its enterprise group, which existed to help other insurers automate operations and fight fraud. He arrived at the AI-for-insurance thesis not from a lab but from the back office, which is a different and more useful place to start. Before Metromile he spent sixteen-plus years in enterprise product, consulting, and engineering, and he holds an MBA from NYU Stern and a master's from Carnegie Mellon.
He is joined by Ryan Eldridge, VP of Engineering and another Metromile alumnus, and Jason St. Pierre, the Chief Product Officer, who previously worked at Twitter, Google, and Verily, Alphabet's life-sciences arm. It is a team that pairs insurance-operations scar tissue with consumer-scale product and engineering - which is roughly the combination you would design if your goal were to make regulated, phone-heavy workflows disappear.
The best-run insurers and agencies run on Liberate's AI.
≈ 130× in about twelve months.
The $50M Series B was all-equity and led by Battery Ventures, whose general partner Marcus Ryu - co-founder and former CEO of Guidewire, the core software many insurers still run on - took a board seat. When the person who built insurance's system of record bets on the AI meant to sit on top of it, that is a signal worth noting. Total funding stands at roughly $72M.
| Round | Amount | Date | Lead & notable investors |
|---|---|---|---|
| Series B | $50M | Oct 2025 | Battery Ventures (lead); Canapi, Redpoint, Eclipse, Commerce Ventures |
| Series A | ~$22M | 2024 | Redpoint Ventures (lead); Eclipse, Commerce Ventures |
| Total raised | ~$72M | — | Valuation: $300M post-money |
Ex-Metromile operators Amrish Singh, Ryan Eldridge, and Jason St. Pierre launch the company to automate insurance operations with AI.
Nicole and end-to-end claims automation begin handling real customer interactions for early carriers.
Fresh capital expands the System of Action across sales, servicing, and claims channels.
Battery leads; the company reports 1.3M monthly resolutions and more than $100B in premium processed.
Liberate sells to insurance carriers, MGAs, and independent agencies - the businesses that field the calls and process the claims. Named customers include Branch Insurance, Allied Trust, Indigo Insurance, Signature Insurance Group, GreatFlorida Insurance, and RightSure. The company designs its agents to plug into the core platforms these firms already run on, names like Guidewire, Duck Creek, and Applied Epic, rather than asking anyone to rip and replace. That integration-first posture is a large part of why an insurer would let AI touch a live claim in the first place.
It builds insurance-native AI agents that automate customer interactions and back-office work - quoting policies, filing claims, and updating endorsements - across voice, email, SMS, and digital channels.
It was co-founded in 2022 by Amrish Singh (CEO), Ryan Eldridge (VP of Engineering), and Jason St. Pierre (CPO), several of whom previously worked at Metromile.
About $72 million total, including a $50 million Series B in October 2025 led by Battery Ventures at a $300 million post-money valuation.
Nicole is Liberate's AI voice assistant that handles inbound and outbound insurance calls - selling policies and servicing requests - with multilingual support and roughly 80% resolution.
Insurance carriers, MGAs, and agencies such as Branch Insurance, Allied Trust, Indigo Insurance, GreatFlorida Insurance, and RightSure. The platform has processed over $100B in premium volume.