GWMM.

A high-performance RAG & search API router that unifies multiple search engines (Brave, Exa, etc.) under a single endpoint. Features sub-second API key auth, user-level usage analytics, and a Python SDK for seamless agent integration.

A transparent, high-performance AI inference network designed for privacy-critical applications. Features zero data retention (ZDR), multi-region routing to local GPU nodes with sub-150ms TTFT, and standard OpenAI-compatible API specifications.

AI-powered legal citation verification tool that checks existence, accuracy, and good-law status of every citation in a legal brief. LLM-powered proposition mismatch detection compares brief claims against actual case holdings, with confidence-scored results delivered in under 30 seconds.

A local, single-binary MCP server that gives coding agents current third-party documentation. Indexes Next.js, React, and Vue docsets locally using hybrid search (semantic + BM25) and reciprocal-rank fusion (RRF) to serve local LLM clients over stdio.