An graph-eval framework for LLM's

The new package transforms a user’s plain‑text search phrase into a structured, BM25‑ranked list of PostgreSQL search results. It takes the query text as input, forwards it to an LLM that rewrites the

radx-destiller selects publications relevant to the NIH RADx-rad initiative using LLMs.

Enterprise-grade, cloud-native, serverless full-stack search platform built with Next.js, TypeScript, & Tailwind. Implements a POST-driven, relevance-scored REST API, debounced React UI, schema-validated JSON contracts, in-memory indexing, stateless execution, & edge-optimized deployment on Vercel for low-latency, high-throughput search workloads.