Rowboat: Open-Source AI Coworker with Knowledge Graph Memory

Rowboat is an open-source AI coworker tool designed to turn your work into a continuously updated knowledge graph. It stores data locally on your machine as editable Markdown files, emulating an Obsidian-style knowledge management system. This tool is especially designed to automate the extraction of context from various sources like Gmail and meeting notes, linking them through backlinks, and updating them as new information comes in.

Key Features

• Living Context Graph: Rowboat captures and organizes information such as decisions, commitments, and deadlines from sources like Gmail, Granola, and Fireflies into a local, living knowledge graph.
• Local AI Assistant: It incorporates local shell access and Model Context Protocol (MCP) support which allows it to use your existing context. For example, generating briefs or creating voice memos using services like ElevenLabs.
• Markdown-Based Storage: Information is stored in plain Markdown files with backlinks, fully editable, allowing users to maintain or modify the data manually.
• Integrations: Rowboat connects with external tools and services like search engines, databases, CRMs, and other automations for expanded functionality.
• Background Automation: Users can automate routine tasks, such as drafting email replies or project updates, based on the context accumulated in the knowledge graph.
• Customizable Model: The tool supports different AI model setups, including local models via Ollama or LM Studio, and hosted models with a provided API key.

This system is particularly beneficial for users who need a tool capable of managing and utilizing dense, context-rich data efficiently without losing time or insights due to slow and lossy traditional search methods.