About Knowledge Raven
Why We Built This
Every company has the same problem: knowledge is scattered across Google Drive folders, Confluence spaces, shared documents, and individual desktops. When someone needs an answer, they spend minutes — sometimes hours — searching, asking colleagues, or re-discovering what the organization already knows.
Existing AI solutions promise to fix this, but most rely on classic RAG — retrieval that returns tiny text chunks (often just 50 words) and only a handful of results. That's enough for simple lookups, but it fails the moment a question requires synthesizing knowledge from multiple documents. We built Knowledge Raven because we believed company knowledge deserves better: a system where the AI agent dynamically searches, fetches, and re-searches until it has enough context for a genuinely well-founded answer.
Our Approach
Knowledge Raven is model-agnostic by design. Instead of building yet another chatbot, we provide MCP tools that work with any LLM — Claude, GPT, Gemini, or any other MCP-compatible agent. You're never locked into a single AI provider.
Under the hood, Knowledge Raven uses Agentic RAG: the AI agent doesn't just run a single search query. It dynamically controls the entire retrieval process — searching broadly across your knowledge base, fetching specific document sections, and re-searching with refined queries as needed. Combined with hybrid search, contextual embeddings, and intelligent reranking, this delivers answers that are grounded in your actual company knowledge, not generic guesses.
Founding Team
Pascal Meger
Product & Technology
AI Manager and technical builder. Architected and built all three BirdFlai products. Drives product vision, system architecture, and hands-on development.
Jan Christiansen
Business & Strategy
Product Manager at Airbus and UC Berkeley Haas alumnus. Brings active industry experience alongside BirdFlai, shaping business strategy and go-to-market.
Knowledge Raven is a product of BirdFlai UG (haftungsbeschränkt), based in Norderstedt, Germany.