SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Town is building a personalized AI assistant that learns from and acts on behalf of users across their tools (email, calendar, documents, Slack, etc.). Founded by Jean-Denis Greze (former CTO of Plaid) and Tony Vincent (former Director of Applied AI Product at Google), the company has raised $73M+ from top-tier investors including Andreessen Horowitz, Forerunner Ventures, and First Round Capital.
You will be the first engineer building Town's search and retrieval infrastructure—the foundation that powers the entire product. This is a greenfield opportunity to architect and implement the systems that enable the AI assistant to surface the right context (memories, messages, documents, relationships) at the right moment.
Key responsibilities include:
- Design and build a unified search and retrieval layer combining lexical and semantic search, serving as a single source of truth for the entire product
- Make critical tradeoffs between vector and lexical search, precomputation vs. on-demand fetching, and hybrid retrieval strategies
- Own the durable data model that underpins the assistant's reasoning and context awareness
- Build and operate realtime/streaming and batch pipelines that keep the index fresh and accurate as underlying data changes
- Architect the indexing and storage layer for scale, optimizing for latency, cost, freshness, and completeness
- Lay groundwork for a knowledge graph relating content by meaning across people, companies, projects, and connections
You should have significant hands-on experience with lexical and semantic search (BM25, embeddings, ANN/vector indexes, hybrid retrieval, ranking), large-scale data infrastructure (both realtime and batch), and the ability to reason through latency/cost/freshness tradeoffs. Experience with ranking, knowledge graphs, or retrieval systems for LLMs and agentic systems is a plus. You're a systems thinker comfortable moving from rapid prototype to production scale.
Location: San Francisco, CA, five days per week in-office at the Financial District office.
