Town launched out of stealth in June 2025 with $55 million from Andreessen Horowitz. Its product is a personal AI assistant called a Townie that reads your calendar and email, briefs you on your day, and auto-generates a wiki about your work patterns, priorities, and colleagues within minutes of onboarding. Pricing runs $49 to $59 per month. CEO Jean-Denis Greze previously served as CTO of Plaid for seven years and ran engineering at Dropbox before that.
The real news is what comes next. Greze told Platformer that a team-wide, self-writing company knowledge base is coming out very soon, one that assembles itself from data across every employee's Townie. The privacy risks are explicit: pool the wrong files and the AI might surface salary data by mistake. Greze's five-year prediction is that high-trust small companies adopt unified AI knowledge bases first, and that specially trained models eventually enforce privacy policies without human review. Enterprise adoption comes last.
The full conversation is worth reading for two specific tensions Greze holds together. He describes Townies in language usually reserved for living things, things you nurture rather than configure, while flatly rejecting AI agents as independent economic actors and calling manipulative cancellation UX a dark pattern. He also says Town's enterprise contracts explicitly prohibit employers from reading workers' Townie sessions, and that session data is deleted after 15 days. Those are policy commitments, not just product choices, and they signal how Town intends to compete against broader workplace surveillance tools.
[READ ORIGINAL →]