query-planning

A high-performance spatial query layer for Polars

Cost-based relational query optimizer (CBO) evaluating nested loop, hash join, and sort-merge joins.

Cost-based relational query optimizer (CBO) evaluating nested loop, hash join, and sort-merge joins.

Spanning Tree-based Query Plan Enumeration

AI revision platform for university past papers, with PDF ingestion, hierarchical chunking, vector semantic search, query planning, and citation-grounded RAG study guide generation.

Active web-search RAG workbench with provider routing, source extraction, citation verification, and extractive fallback.

LLM query planner that compiles natural-language questions over 1M+ rows of Boston open data into typed SQL, retrieval, or cross-dataset plans, benchmarked at 86.7% accuracy vs 20% for RAG baselines

🏥 Modular Agentic RAG Medical Assistant 🤖 built with LangGraph, Pinecone, Groq & Tavily. It intelligently routes queries between RAG, live web search, and direct answers. 🔬 Live tracing shows every agent step, while FastAPI + Streamlit deliver a production-ready full-stack experience. 🚀

Agentic pipeline for natural-language question answering over federated RDF graphs, with SPARQL endpoint discovery, schema-aware query planning, validation retries, and SPIDER4FedSPARQL benchmarking.

Intelligence layer for AI agents. Query planning, context discovery, cost optimization. 60-75% token reduction while maintaining quality.