Town Raises $108M Series B for AI That Works Across Teams

Town gives each AI assistant a person to answer to, even when the assistants start working together. That design puts an ordinary office judgment inside the software: knowing something about a colleague's work does not automatically make it yours to pass along.

The San Francisco company announced a $108M Series B on October 19, 2026, co-led by Forerunner Ventures and Menlo Ventures at a $1B valuation. Co-founders Jean-Denis Greze, CEO, and Tony Vincent, CPO, are extending their personalized assistants, called Townies, into coordinated work across teams. The financing follows a $55M Series A announced in June.

Town's opportunity sits in the exchanges between people: a question answered, a draft reviewed, a process improved and carried into somebody else's day. Its expansion gives enterprise AI buyers a concrete reason to examine how personal context, company ownership and permission fit together before delegating more work.

What Town's Series B is financing

Town's founders describe a product that can connect one person's work with another's without requiring both people to manage every exchange. Their funding essay describes assistants finding warm introductions, gathering feedback before a meeting and applying a process change across a team. These are product capabilities and ambitions described by Town, rather than independently measured productivity results.

The useful detail is the handoff. An assistant preparing a message for an executive needs to understand the proposed communication and the executive's preferences. A teammate's assistant may hold part of that context. Town is building the relationship through which the information can move, while leaving consequential human decisions with the people responsible for them.

Menlo is arriving with experience as a customer. In its investment account, the firm says it deployed Town across the business in June and used assistants for inbox triage, briefings and follow-ups. That account explains the firm's enthusiasm; it is also a testimonial from a new investor, with the interests that entails.

Existing investors Andreessen Horowitz and First Round Capital joined the financing. Town also names Better Tomorrow Ventures, Precedent.vc, Scribble Ventures, Serena Ventures and Sparked Ventures, alongside individual angels. The announced amount is new Series B funding, while the $1B figure describes valuation. Neither number tells a buyer how much work a Townie can reliably take off a particular team's plate.

Personal context becomes a team resource

Town's product development gives the team announcement some history. A September update already described Townies participating in group texts, asking other assistants for information and using a browser with approval at relevant steps. The October positioning develops that collaboration into an explicit company-wide proposition. It would be misleading to describe every assistant-to-assistant interaction as newly introduced with this round.

The personal side has its own machinery. In July, Town described the Town Wiki, a user-editable account of the assistant's understanding of a person's profile, goals and projects. The company says a nightly process consolidates memories, removes stale information and checks what has changed. A useful assistant needs some way to update its understanding when the person it represents changes direction.

Teamwork adds a distribution problem to that memory problem. Town's example is a change to competitive battle cards that can flow into every sales representative's call preparation. The operational appeal is easy to understand: improving the shared process should improve the next customer conversation without making every representative maintain a separate set of instructions.

The same buying concern appears in DevCuration's coverage of Opsin's agent-security financing. For a buyer, this changes the evaluation. The demo can show that an assistant finds a useful answer. A team deployment also needs to show whose information supported it, how a change propagates and where a person can intervene. Those questions follow directly from the collaborative design Town is selling.

Founders, early growth and the capital history

Greze and Vincent have worked together since Dropbox. Town's current team page identifies Greze as the former CTO of Plaid and Vincent as the former Director of Applied AI Product at Google. Their Town roles are CEO and CPO, respectively. That mix of infrastructure and product experience is relevant to a company asking users to connect sensitive working systems while making the experience feel approachable.

Town's June 3 announcement disclosed a $55M Series A led by a16z. The two disclosed rounds sum to $163M, but the releases do not establish a complete lifetime financing history. Treating a rounded total from a secondary report as an exact capital ledger would add certainty the evidence does not supply.

The October release says Town reached seven-figure annual recurring revenue within weeks of its June public launch. It also reports that paying customers grew 4x and paying teams grew 9x since July. Those are company-reported figures; absolute customer counts, retention and audited revenue are not supplied alongside them.

The relative team growth matters to the strategy because colleagues become part of the experience. A user may value an assistant for individual preparation, then find another use when a colleague's assistant can help resolve a shared task. How durable that behavior becomes will emerge in customer relationships and renewals, beyond the early growth ratios available today.

Approval is part of what companies are buying

Town's trust policy says the service requires approval before sending emails or calendar invitations. It also says model training is off by default, with an opt-in for eligible data, and that Google-connected data is never used for training. These are specific company policy representations. They should be read alongside the permissions and contractual terms of a proposed deployment.

The company says it is SOC 2 Type 2 audited, with reports available under an NDA, and publishes a data processing agreement. The private audit was not reviewed for this article. Town also says companies own the context their assistants develop and can take it with them if they leave, giving knowledge ownership a place in the commercial proposition. It is a question that also runs through RapidCanvas's enterprise-context work.

For enterprise AI, that makes the working relationship unusually visible. Each connected inbox, shared routine and delegated conversation can become useful to several people, while responsibility still has to land somewhere. A sales leader buying coordinated call preparation and an employee connecting private correspondence are participating in the same product through different expectations.

Town now has substantial backing to expand that arrangement. As more colleagues bring their assistants into a shared task, the practical question keeps following the information: what can this Townie pass along, and whose decision does it still need?