Interpreter
Foundational Model for Parts
Interpreter is an AI model purpose-built to handle the complexities of parts. This is the breakthrough technology the industry has been waiting for.
.png)
Built by automotive veterans and world-class engineers from companies like:
How it’s built
Architecture delivering accuracy at scale
Interpreter is built on a multi-modal architecture, processing technical diagrams, damage photos, and text descriptions. Training data comes from verified sources and is annotated by experienced parts interpreters.
Curated training data
Partly sources from government records and licensed manufacturer data, and conducts original research through structured vehicle tear-downs and expert analysis.
Purpose-built annotation tools
Partly developed annotation software in-house to handle the unique complexities of parts data that generic labeling tools can't address.
Expert labelling
Training data is annotated by experienced parts interpreters and verified against deployed production data.
Interpreter delivers standardised, precise and complete vehicle and parts information.
9x
Faster order processing
2.7x
Reduction in supplementaries
2.4x
Less parts returns and orders cancelled
Standardised
Universal standard across manufacturers
Every manufacturer structures their parts catalogs differently. Interpreter delivers a single standard for how vehicles break down into assemblies, how parts within those assemblies are grouped, and naming conventions across the industry.
Precise
Exact right part, every time
Interpreter delivers the exact parts for a VIN. The model accounts for vehicle variants like trim level, engine type, market, and production date to determine fitment.
Complete
Complete context for every job
Interpreter provides complete parts information beyond what's in most manufacturer catalogs. This includes installation hardware requirements, market-specific modifications, part supersessions, and finish specifications.
AI is powerful, proven and scalable.
