Rowboat is an open-source AI coworker written in TypeScript that maintains a persistent knowledge graph of your work context to help you accomplish tasks more efficiently. The core innovation is that instead of reconstructing context on demand through searches, Rowboat builds and maintains long-lived memory stored as plain Markdown notes in an Obsidian-compatible vault that lives entirely on your machine. This approach allows context to accumulate over time with explicit, inspectable relationships between people, projects, decisions, and commitments.
The application connects to your existing work tools including Gmail, Google Calendar, and meeting notes from Rowboat's own note-taking system or Fireflies. It automatically extracts and organizes relevant information into your knowledge graph, which you can visualize, edit, and update at any time. The system is designed to help with concrete work tasks like generating presentation decks about quarterly roadmaps, preparing meeting briefs that pull together past decisions and open questions, tracking people or companies through live notes, and recording voice memos that automatically update key takeaways in the graph.
Rowboat's architecture emphasizes local-first operation and user control. All data is stored locally as plain Markdown files with no proprietary formats or hosted lock-in, meaning users can inspect, edit, back up, or delete everything at any time. The application supports flexible model configuration, allowing users to bring their own models through local options like Ollama or LM Studio, or use hosted models with personal API keys. Users can swap models anytime without affecting their locally stored data.
The platform extends functionality through Model Context Protocol integrations, enabling connections to external tools and services like Exa for web search, ElevenLabs for voice capabilities, Slack, Linear, Jira, GitHub, and custom internal tools. Optional integrations include Deepgram for voice input, ElevenLabs for voice output, and Exa for research search capabilities. The application is available as downloadable binaries for Mac, Windows, and Linux.
According to GitGenius activity tracking, the repository shows median issue and pull request response latency of 7.9 hours across 111 tracked items, with mean latency of 591.3 hours. The most active contributors tracked are arkml with 68 events, ramnique with 46 events, and tusharmagar with 13 events. The project overlaps with contributors from related repositories including yeachan-heo/oh-my-claudecode, microsoft/playwright, and langgenius/dify. The repository is classified across multiple data-focused categories including data orchestration, data pipelines, workflow management, data engineering, data platform, data quality, data lineage, observability, scheduling, and Kubernetes. Rowboat is a Y Combinator S24 company and maintains active community channels on Discord and Twitter.
