Search for parts
the way a workshop
actually talks.
A domain-specific embedding model trained on India's auto-part catalogs, OEM cross-references, and the Hinglish a mechanic actually uses. Because "Swift dicky strut" should find the tailgate gas spring — not nothing at all.
A buyer asks for a "Scorpio dicky wala strut, 2015, diesel ka". The catalog lists it as Tailgate Lift Support · 0609AA0044N. Keyword search misses. Fuzzy match misses worse.
General-purpose embedding models (OpenAI, Cohere, Gemini) were trained on the open web. They don't know that a dicky is a boot, a silencer is an exhaust muffler, or that Alto K10 parts interchange with Maruti 800 on 60% of SKUs. PartIndex is trained on that language.
Live against a 1.2 M-part index.
A compact bi-encoder, fine-tuned on triplets of query, matching part, and near-miss — with a vehicle-compatibility graph underneath.
260M parameters, runs on a single A10 at under 50 ms p95. No LLM round-trip in the hot path. Every SKU is bound to the vehicles it fits — so "will this clutch work on my Scorpio?" gets a straight yes or no, not a confidence score dressed up as one.
Shipping to ten partners in Q2 2026.
If you run an auto-parts marketplace, a distributor network, a workshop-facing app, or an OEM e-commerce surface — one line about your use case is enough.