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.
{ "caseStudies": [...] }
// filter by tag:
all
agents
computer-vision
elasticsearch
iit-kanpur
innocean
legal-tech
martech
ml
ocr
rag
AdINNtelligence: what should our next ad be?
Competitive ad intelligence for car dealerships: collect competitor Meta ads, classify them with LLMs, and return three evidence-backed recommendations.
ScoutINNtelligence: supply scan and demand scan
Dual-intelligence lead generation: rank startups by fit on one side, qualify buyers with a 3-filter check on the other.
Law Simplified: making legal documents readable
OCR, clause and entity extraction, schema-driven summaries, and a chat interface that answers only from the uploaded document.
Evidence Analysis: telling site photos from documents
A multi-factor scoring and voting pipeline for Kanpur Smart City that catches documents uploaded where a site photo was expected.
Classifying drones from radar cross-section
Multi-frequency (26–40 GHz) RCS measurements, cubic spline interpolation across phi–theta angles, and 96.6% accuracy on drone type.