Verified project record
SenteLabsAI/OpenExecutive
OpenExecutive is a self-hostable, open-source AI-powered virtual executive that answers in one consistent executive voice backed by a council of specialist agents. It grounds responses in built-in MBA-level knowledge plus the company's own documents, with episodic memory, a proactive scheduler, and chat-app integrations.
Project overview
It targets a specific gap—AI assistance for leadership and management work that lacks company context, governance over colleague interactions, and spend-approval rules—by combining a two-layer RAG knowledge base, a guided company onboarding wizard, and integrations with Discord, Telegram, and MCP clients, while remaining self-hostable with optional local or OpenAI-compatible model routing.
- Project type
- MCP · AI Agent · RAG
- Use cases
- Knowledge Q&A · Automation
- Deployment
- Refer to project documentation
- License
- License pending
Best for
- Enterprise teams that want leadership and management assistance grounded in their own company documents and profile, delivered in one consistent executive voice.
- Developers and AI engineers who need to self-host and choose their inference endpoint, including local servers such as Ollama, LM Studio, vLLM, or llama.cpp, or OpenAI-compatible gateways.
- Teams already using MCP clients (Claude Code, Claude Desktop, Cursor) that want to connect to the embedded /mcp server.
Key capabilities
- Built-in MBA-level Markdown knowledge seeded into ChromaDB plus user-uploaded company documents chunked into a company_docs collection, with RAG context injected into the user turn.
- Built-in job runner surfaces follow-ups and time-sensitive actions automatically.
- Guided wizard collects a company profile that the Executive references in every response.
- Can run against local servers (Ollama, LM Studio, vLLM, llama.cpp) or hosted OpenAI-compatible gateways instead of the Anthropic API.
- Executive and specialists can answer with live web results (news, market data, competitor moves) via Anthropic's server-side web_search tool.
Limitations and risks
- Server-side web search, Anthropic prompt caching, and extended thinking are automatically disabled for local models.
- Company profile, documents, and conversations are sent to whichever LLM endpoint is configured; before deployment, verify settings point only at a trusted provider.
- Multi-agent routing leans heavily on tool use, and small local models may route poorly.
Getting started
- Setup is medium difficulty: it requires cloning the repo, copying .env.example, and setting an API key; the first run pulls heavy ML dependencies and downloads a ~90 MB embedding model, taking a few minutes.
- git clone https://github.com/SenteLabsAI/OpenExecutive.git && cd OpenExecutive, then make install, then cp .env.example .env and set ANTHROPIC_API_KEY (or local/OpenRouter config), then make dev, then open http://localhost:3000 and complete onboarding.
Evidence and sources
- GitHub project description: AI-powered virtual executive team — a single coherent executive persona backed by 8 specialist agents (FastAPI + Next.js).
- README: Open Executive is designed to transform leadership and management. Highly configurable, it can be deployed at any management level. Out of the box it supports spend approval thres…
- README: All responses come from one consistent executive voice. The internal agent architecture is never exposed to the user.
- README: Two retrieval layers per specialist call: (1) built-in MBA-level Markdown (`knowledge/builtin/`, git-tracked) seeded into ChromaDB at startup, and (2) your uploaded company docume…
- README: Beyond Q&A, the system maintains episodic memory of past decisions and initiatives across sessions, and a built-in scheduler can proactively surface follow-ups and time-sensitive…
AI Search
Find projects, verify facts, compare options, or turn a complex need into an actionable plan