Aurex'26 · Track 05 — Universal Knowledge Assistant

THAMIZHARIVU AI

Ask in Tamil. Discover the Knowledge. Verify the Source.

A Tamil-first, source-grounded knowledge discovery system for classical literature, cultural heritage and digital archives. Every answer carries its evidence: source title, type, link and the passage it came from.

Tamil palm-leaf manuscripts beside a Dravidian temple gopuram with kolam patterns

Why it is different

Not a generic chatbotவேறுபாடு

A knowledge discovery system built around retrieval, ranking and citation — not free-form generation.

Understands how you speak

Tamil, English, Tanglish, transliteration and mixed queries — no fixed question lists or keyword rules.

Discovers sources dynamically

Webpages, books, e-books, PDFs and digital archives are discovered, retrieved and ranked per query.

Never invents facts

Answers stay grounded in retrieved passages. Unverifiable claims are flagged, not guessed.

Ready for the live demo walkthrough

The interface layer is complete and wired for the retrieval pipeline: LLM, RAG, embeddings, vector search, hybrid ranking, conversation memory and Tamil OCR arrive in the next phases.