Open Source Hacker News (AI)

CEO fired developers to make room for AI. Developers create open source AI CEO

Open ExecutiveAI agentsopen sourceAnthropic Claude

Open Executive, developed by sentelabs.ai, is an open-source AI system that acts as a virtual executive team, providing a single coherent executive voice backed by eight specialist agents: Chief Strategy Officer, Chief Financial Officer, Chief HR/People Officer, General Counsel, Chief Operating Officer, Chief Marketing Officer, Chief Product Officer, and Board Communications Director. The internal agent architecture is hidden from the user, and all responses come from one consistent voice.

The system routes user messages through an Executive Orchestrator (claude-sonnet-4-6) that makes parallel tool calls to the specialists, each retrieving relevant context from ChromaDB. Knowledge comes from two retrieval layers: built-in MBA-level Markdown and user-uploaded company documents. It also maintains episodic memory in SQLite, with a background claude-haiku-4-5 pass extracting key decisions and initiatives after every response, injected as a <past_decisions> block in future sessions. A built-in scheduler uses UPDATE…RETURNING to prevent double-firing and requires running as a single instance. Prompt caching is structured to achieve up to 85% cache hit rates by separating the executive persona, company profile, and knowledge index.

The tech stack includes Python 3.11 + FastAPI, uv, ChromaDB, SQLite, Next.js 15, Tailwind, and Anthropic Claude models (default claude-sonnet-4-6, with claude-opus-4-7 for deep reasoning on CSO, CFO, GC, Board). The project is released under Apache 2.0.

This project turns the 'AI replaces workers' narrative on its head, offering a free, open-source alternative for companies to experiment with AI-driven executive advice. Its modular architecture and memory features make it a notable community contribution.

Read original →

← Back to home