ES-RAG: natural-language search over government grievances
An LLM turns plain-English queries into structured Elasticsearch queries. A semantic cache and hybrid retrieval sit behind it, and a RAG layer answers.
An LLM turns plain-English queries into structured Elasticsearch queries. A semantic cache and hybrid retrieval sit behind it, and a RAG layer answers.
Competitive ad intelligence for car dealerships: collect competitor Meta ads, classify them with LLMs, and return three evidence-backed recommendations.
Dual-intelligence lead generation: rank startups by fit on one side, qualify buyers with a 3-filter check on the other.
OCR, clause and entity extraction, schema-driven summaries, and a chat interface that answers only from the uploaded document.
A multi-factor scoring and voting pipeline for Kanpur Smart City that catches documents uploaded where a site photo was expected.
Matching a single radar reading against spline-fitted radar cross-section signatures of 17 drones and objects, with a phi/theta vote. 96.6% accuracy on drone type.
This is the plain-HTML version of this site — no JavaScript, CSS 2.1 only,
built to render in Dillo.
Full version of this page · GitHub · LinkedIn · Resume · ashwin.mishra07@gmail.com