Good product, but "100% deterministic" is marketing—hard stop on the Fact engine (TF-IDF for fact verification), Reasoning engine, and Image verification. The Math and Logic layers using SymPy/Z3 are genuinely deterministic. But claiming TF-IDF can verify facts deterministically is just keyword-document similarity—it'll miss context, evolving information, and non-obvious logical chains. What you're really building is a strong gate for verifiable domains (math, structured logic, code security) layered on top of weaker heuristics for fuzzy domains. That's honest work and useful. The positioning should match: "deterministic where computable, high-confidence verification elsewhere." As it stands, teams will trust the whole output with 100% confidence when only parts earn it.

I built QWED – a verification layer that sits between your LLM and production. The idea: Don't fix hallucinations, verify them. If AI output can't be mathematically proven, it doesn't ship. 11 specialized engines: - Math (SymPy) – verify calculations - Logic (Z3 SMT) – formal proofs - SQL (SQLGlot) – detect injection/dangerous queries - Code (AST) – security analysis + taint tracking - Facts (KB) – entity verification without LLM Works with ANY LLM – OpenAI, Claude, Gemini, or local models via Ollama ($0). Model-agnostic: Your LLM choice, our verification. Happy to answer questions about deterministic AI verification!