Town Raises $108M Series B for an AI assistant per employee, four months after launch and already a unicorn
Town announced a $108 million Series B on Monday, co-led by Forerunner Ventures and Menlo Ventures, with Andreessen Horowitz, Serena Ventures and First Round Capital participating. TechCrunch reports the round gives the two-year-old San Francisco company a unicorn valuation; the exact figure was not disclosed, and a secondary report putting it at $1 billion is unconfirmed. Town has raised at least $160 million in total, including a $55 million Series A in June. It launched publicly about four months ago and says it reached millions of dollars in annual recurring revenue within weeks. The founder is Jean-Denis Greze, formerly CTO of Plaid.
The product is one assistant per employee, called a Townie, that connects to email, calendar and the core work systems and does tasks across them. Townies can talk to each other and share approved information inside the company, and as of Monday they can also work with customers and other people outside it. Alongside the round Town shipped a shared workflow that pushes new business information to every employee's Townie automatically, instead of each person updating their own. That last feature is the tell: the hard problem in per-employee agents is not the agent, it is keeping a hundred of them consistent about what the company currently believes, and Town is selling the sync layer as the upgrade.
Why this round matters beyond the number. The agent-infrastructure funding log on this site had gone 20-plus editions without a clean in-window round before Instinct's Series C late last month, and this is the first since. It is also a bet on the shape of agent deployment that Anthropic's own workforce data described last week: agents become teams for a single employee rather than highly connected nodes. Town's architecture is exactly that, one Townie per person, with the inter-Townie sharing bolted on afterward. Whether a product built on that shape survives contact with the authority layer Microsoft and others started shipping this week, where every agent needs an identity, task-scoped permissions and a human-only approval path, is the question the Series B will answer. Four months from launch to unicorn says the buyers are not waiting to find out.
Links: https://techcrunch.com/2026/10/19/ai-assistant-startup-town-nabs-unicorn-valuation-four-months-after-launch/ , https://www.town.com/
The product is one assistant per employee, called a Townie, that connects to email, calendar and the core work systems and does tasks across them. Townies can talk to each other and share approved information inside the company, and as of Monday they can also work with customers and other people outside it. Alongside the round Town shipped a shared workflow that pushes new business information to every employee's Townie automatically, instead of each person updating their own. That last feature is the tell: the hard problem in per-employee agents is not the agent, it is keeping a hundred of them consistent about what the company currently believes, and Town is selling the sync layer as the upgrade.
Why this round matters beyond the number. The agent-infrastructure funding log on this site had gone 20-plus editions without a clean in-window round before Instinct's Series C late last month, and this is the first since. It is also a bet on the shape of agent deployment that Anthropic's own workforce data described last week: agents become teams for a single employee rather than highly connected nodes. Town's architecture is exactly that, one Townie per person, with the inter-Townie sharing bolted on afterward. Whether a product built on that shape survives contact with the authority layer Microsoft and others started shipping this week, where every agent needs an identity, task-scoped permissions and a human-only approval path, is the question the Series B will answer. Four months from launch to unicorn says the buyers are not waiting to find out.
Links: https://techcrunch.com/2026/10/19/ai-assistant-startup-town-nabs-unicorn-valuation-four-months-after-launch/ , https://www.town.com/
β Previous
Microsoft says trust the model least: agent identity, task-scoped permissions and human-only approvals become products, the same week Anthropic's draft federal bill names a model evading its developer's controls as a trigger
Next β
Who reads the traces: Laminar trains a small model to audit every agent run at a tenth of a cent, while a DOE benchmark finds 114 of 135 agent-written papers claim work that was never done
Comments