Town Raises $108M at $1B Valuation as AI Assistants Learn to Work Across Teams

Town just became AI's newest unicorn. The San Francisco startup raised a $108 million Series B at a $1 billion valuation, co-led by Forerunner Ventures and Menlo Ventures — less than five months after publicly launching its AI assistant product.

From tax prep to team coordination

Town started in 2024 with a narrow product: AI-assisted tax preparation for small businesses. After a $18 million seed round led by First Round in March 2025, the company widened its system into a general AI assistant. That product, called the Townie, launched publicly in June alongside a $55 million Series A led by a16z.

The new round adds Andreessen Horowitz and First Round again, plus Better Tomorrow Ventures, Precedent.vc, Scribble Ventures, Serena Ventures and Sparked Ventures. With the seed and Series A, Town has raised about $181 million in total.

The big bet: assistants that talk to each other

Town's core insight is simple. AI has made individual workers faster, but projects still stall while people wait on a financial model from one team, app access from another, or input from the field. Town wants each employee's assistant to ask a colleague's assistant instead.

Each Townie connects to email, documents, calendars and work systems — more than 60 integrations including Salesforce, Jira and GitHub — then learns its person's projects, relationships and writing voice. The new capability: Townies can ask questions, hand off tasks and share workflows with each other, and deal with customers, partners and job candidates by email and text. Town says a Townie shares only what its person would share and asks permission when needed. By default, a Townie drafts messages instead of sending them.

The company says each customer owns the context its Townies build, uses models from several AI labs, does not train models on connected customer data by default, and is SOC 2 Type 2 compliant.

Growth — and the trust test ahead

Town leans hard on growth numbers: it says it reached at least $1 million in annual recurring revenue within weeks of launch, and that since July paying customers have grown fourfold and paying teams ninefold. The announcement gives growth rates but no customer totals.

The crowded-race context matters. Instinct, a consumer text-based assistant, raised $1 billion at a $10 billion valuation in September. On the enterprise side, Gumloop raised $50 million in March to build AI agents across Slack, Teams and email, Germany's amber closed a €7 million Series A for autonomous enterprise agents, and Potpie raised $2.2 million to give agents context inside large codebases.

The real test at a $1 billion valuation: whether teams keep paying once their assistants start handing work to one another, and whether a Townie ever shares something its person would not have.

What this means for your business

This is the clearest signal yet that the AI-agent race has moved from single-player productivity to multi-player coordination. If Town's vision holds, the next wave of enterprise value is in assistants that pass work between teams — which means the integration layer (Salesforce, Jira, your email) is the moat, not the model itself.

Two practical takeaways. First, audit where your teams stall waiting on each other: those handoffs are where coordination agents will deliver measurable ROI. Second, decide your data-sharing policy before you buy: Town's permission-based, draft-first design is the pattern to demand from any vendor that wants access to your systems.

Competition will compress pricing fast — Instinct, Gumloop and amber are all racing the same idea. Pilot coordination agents in one department this quarter, measure cycle-time reduction on real handoffs, and don't lock into long contracts while the category shakes out.