Town

Town is an AI work assistant that learns from your existing work to automate routine tasks. It analyzes your emails, calendar, and documents to understand your clients, deadlines, and writing style, eliminating the need for endless prompt engineering. Designed for busy professionals and teams, Town offers a zero-setup experience where the assistant simply observes and then acts on your behalf. The core value is saving hours each week on repetitive administrative work while maintaining your personal voice and preferences. Rather than requiring users to manually teach it every nuance of their workflow, Town silently ingests your existing data—the emails you write, the meetings you attend, the documents you create—and builds a model of your habits. This means your assistant already knows how to handle your most common tasks from day one, just like the assistant 'Bo' did for Jess in the provided examples.

The concrete problem Town solves is the significant time drain caused by manual, repetitive tasks such as drafting routine emails, preparing for meetings, and monitoring project updates. Professionals often spend hours each week on these activities, which detract from high-value strategic work. Additionally, traditional AI assistants require extensive upfront configuration and ongoing prompt refinement, creating a learning curve that many never fully overcome. Town eliminates this friction by learning directly from your existing workflow. For example, the assistant 'Snap' automatically checks Beth's client threads every Friday and writes status summaries, so Beth wakes up to completed reports. This matters because it frees up mental energy and time, allowing users to focus on their core responsibilities without the overhead of managing an AI tool.

First major feature group is 'Routines,' which are pre-built or custom automations that handle recurring tasks. One notable routine is 'Auto-inbox,' which watches for new emails, applies labels based on message type, drafts responses when a reply is clearly warranted, and helps with scheduling. Another is 'Meeting Briefing,' which researches upcoming meetings, skips low-signal events, emails you a concise briefing, saves it as a Town document, and can optionally add a private calendar copy. These features work by integrating directly with your email and calendar, then executing on a set schedule or trigger. The benefit is that users no longer need to manually sort through their inbox or prepare meeting materials; the assistant handles the heavy lifting, ensuring nothing falls through the cracks.

Second major feature group is the assistant's ability to learn from your existing work without any prompts. As demonstrated with Jess, the assistant 'Bo' learned her clients, deadlines, and the way she writes simply by observing her previous work. This is a fundamental departure from traditional AI tools that require users to define every parameter. Town's assistant ingests your emails, documents, and calendar events to understand your language, priorities, and patterns. It then applies this knowledge to take actions that match your style, such as drafting emails in your voice or prioritizing tasks based on your habits. This eliminates the step of teaching the assistant how you work, as noted by investor Athena Shiravi.