Town AI (Townie): What It Does, Pricing, and Alternatives for Teams (2026)

TL;DR

• Town AI is a work assistant from Town (town.com) that gives each employee a personal agent called a Townie. Townies connect to their person's email, docs, calendar and core work systems, and on a team they ask each other questions and hand off work.
• Town is a fast-growing product: a public launch in June 2026, then a $108 million Series B at a $1 billion valuation on 19 October 2026. Plans run from free to $199 per user per month, with Team and Enterprise tiers.
• Its design is per person. Each Townie works for its own person and shares only what that person would share, so the company's picture is assembled one assistant at a time.
• Sentra is the company memory Townies and other agents can check with, joining ERP, CRM and HRIS with what people said in email, chat and meetings, with a named person approving before anything goes out.

What is Town AI?

Town AI is a workplace assistant from the company Town that gives every person at work their own AI agent, a Townie, which learns how they work and then handles email, scheduling, meeting prep, documents and recurring routines across their tools (town.com). Town's tagline is "Learns how you work, then gets to work" (town.com).

If you searched "town ai" for something else: AI Town is a separate, MIT-licensed open-source starter kit from a16z for building a virtual town where AI characters live and interact (github.com). This page is about Town, the Townie work assistant.

Town announced its public launch in June 2026 alongside a $55 million Series A led by a16z. On 19 October 2026 it announced a $108 million Series B at a $1 billion valuation, co-led by Forerunner Ventures and Menlo Ventures, and said paying customers had grown 4x and paying teams 9x since July (town.com, town.com).

What a Townie does

A Townie connects to its person's email, docs, calendar and the core systems they work in (town.com). Town's product page describes it identifying what matters in an inbox, drafting replies in your voice, catching loose ends and joining email threads, and using its own browser for work beyond integrations, within the permissions you set (town.com).

• Routines. Hand off repeated work, run tasks on a schedule or when something new arrives. Examples include morning briefings, meeting briefings, inbox triage, deal spotting and competitive intel briefings (town.com).
• Finance chores. Town's examples include drafting reminders for invoices 30 days overdue in QuickBooks Online and matching receipts to Ramp charges (town.com).
• A knowledge base on how you work. Each Townie keeps a wiki of what it knows about your people, projects and priorities, learned from connected tools, feedback and corrected drafts (town.com).
• Approvals. "By default, your Townie asks before anything important goes out" (town.com). Anything that sends, books, shares or writes to another system waits for permission, set per tool as "Ask each time" or "Always allow" (town.com).
• Connectors. Gmail, Outlook, Slack, Google Calendar and Workspace apps, HubSpot, Salesforce, Notion, Asana, Linear, GitHub, QuickBooks Online, Brex, Ramp, iMessage and more, plus "any MCP" (town.com).

Town for teams: how Townies work together

Town's team model is one assistant per person. "Everyone gets their own Townie, not one bot shared by the whole channel," and each one knows its person's work, so together they fill gaps and keep things moving (town.com). The Series B announcement puts it this way: Townies "ask each other questions, connect the dots, and hand off work," and each works for its own person, sharing only what that person would share (town.com).

On a team you can add a teammate to a Town thread, group text or Slack channel and their Townie joins. Teams bundle routines, skills and integrations into a "Power" that teammates' Townies set up with their own accounts, and updating a shared process once updates every Townie (town.com).

Town AI pricing

Town publishes its individual prices (town.com).

Team plans pool credits: Team includes 6,000 pooled credits per seat and Team Plus 18,000, with per-seat pricing that the pricing page does not list. Enterprise adds unlimited seats, custom rates, dedicated support and a security review. Overage is billed per credit, and every account has an overage cap that defaults to $0 (town.com).

On security, Town says it is SOC 2 Type 2 compliant, publishes a DPA, and does not train models on what customers connect by default (town.com).

What Town AI can see, and what it can't

Each Townie sees its own person's accounts. A Townie connects to its person's email, docs and calendar (town.com), and you decide what each Townie can see and share (town.com). That keeps permissions simple. It also means the company's knowledge is split across as many assistants as there are employees.

Company knowledge travels by asking. When one Townie needs something another person knows, it asks that person's Townie, which shares only what its person would share. Facts that live with someone who is on leave, has left, or never connected a tool are not reachable this way. Town's documentation describes each Townie's wiki of "your people, projects, and priorities"; it does not describe a single company-wide record that all Townies read from (town.com).

Every Townie starts from its person. A Townie learns from connected tools, feedback and examples of good work (town.com). A new hire's Townie starts with that new hire's accounts, not with the history of the accounts and projects they inherit.

Connected apps are separate sources. Town reads QuickBooks, Ramp, Salesforce and HubSpot through connectors. Its pages do not describe reconciling a system of record with the conversations around it, for example tying a revenue shortfall in the ledger to the pricing decision made on a call.

For a team of ten, these are small costs against real time saved. They grow with headcount, which is the problem our guide to personal AI agents at work explores.

Works with Sentra

Townies know their people. Sentra knows the company, and every Townie can ask it.

Sentra reads what is happening across the company, in email, chat, meetings and documents and in ERP, CRM, HRIS and BI, and works out what is true now, why, who is working on what and toward which goals. It knows when a fact became true and when it stopped being true, so an assistant asking who owns an account after a reorg gets today's owner. Answers respect each source's existing permissions, checked at query time, and a failed check counts as no access, so a Townie acting for one person sees only what that person may see.

Because the memory is shared, it does not depend on the right colleague's assistant being available. Decisions and commitments from people who left are still there, with when they were made and whether they still hold.

Sentra exposes that memory over REST and MCP, and Town says Townies work with "any MCP" (town.com). The same memory serves ChatGPT, Claude, Copilot, Gemini and internal agents.

Sentra also acts first: it finds what is going wrong, works out who should act and tells them, and a named person approves anything before it reaches anyone else. The Sentra Agent catches promises made in calls and threads and texts people before they are due on iMessage, WhatsApp or Slack.

Sentra is SOC 2 Type II and ISO 27001 certified, connects 200+ integrations, and deploys in the cloud, in an isolated VPC, or air-gapped on premises. It replaces none of your tools.

Book a demo at /demo to see Sentra under Town and the other agents your team uses.

Town AI alternatives

Sources: learn.chatgpt.com, support.claude.com, sauna.ai.

For a wider view, see AI coworker platforms and Sentra vs Claude Tag.

How to choose

1. Pick the unit of work. If you want each person's inbox and calendar handled, a per-person assistant such as Town fits. If you want what slips between people caught, you need a view across the company.

2. Count the people who hold key knowledge. The more your answers depend on specific colleagues, the more a per-person model depends on their assistants being connected and willing to share.

3. List your systems of record. Ask each vendor how its agent joins ERP, CRM and HRIS records with the conversations that explain them.

4. Check the price at your team size. Town's individual tiers run from $15 to $199 a month; get Team and Enterprise quotes in writing.

5. Plan for several assistants. Employees will use Town, ChatGPT and Claude side by side. One shared memory under all of them keeps their answers consistent.

FAQ