JOY Automart · Automotive platform
Automotive Intelligence
JOY Automotive Intelligence organises vehicle identity, symptoms, diagnostic evidence, compatible-part context and sourcing options into one reasoning surface, so a vehicle problem becomes a structured next action instead of an open search box.
Vehicle identity comes first
VIN and chassis context, model year, engine, drivetrain, market specification, option codes, OEM references and part supersessions all change what actually fits a vehicle. JOY treats these as required inputs rather than optional filters.
Where exact configuration evidence is unavailable, the correct machine-readable answer is “verification required”. Automated clients must not infer guaranteed compatibility from a generic product title alone.
Diagnostics are evidence, not verdicts
Diagnostic Trouble Codes and reported symptoms are evidence inputs. A complete repair decision may additionally require scan data, freeze-frame information, live measurements, service history, physical inspection, wiring and power checks, published technical procedures and post-repair verification.
JOY separates a hypothesis from a verified fault throughout the interface and the API, so no surface represents an unconfirmed suggestion as a completed repair.
HERMES and MAYA
HERMES is the JOY agent framework. MAYA is the public AI receptionist and automotive assistant. Public AI access never exposes Owner or Admin authority.
A response from an AI surface is advisory unless the product explicitly returns execution evidence for an authorised action. A model recommendation is not proof that an external action occurred.