A working technology that connects automotive workshop documentation to an AI assistant — enabling natural language search, multi-platform diagnostic data log interpretation, and end-to-end diagnostic workflows that do not exist anywhere else in the market.
Three steps that replace hours of manual documentation search and years of diagnostic experience.
Type a question in plain English — a DTC code, a drivability complaint, a fluid spec. The system searches the complete workshop documentation library instantly using semantic understanding, not keyword matching.
The system identifies the exact PIDs to capture for your scan tool of choice -- FORScan, Torque Pro, OBD Fusion, HP Tuners, dealer-level IDS or FDRS, or any other OBD2 platform -- the drive conditions to log under, and the normal vs. abnormal thresholds for each parameter on that specific engine.
Share the data log. Every variable is explained, cross-referenced against workshop specifications, and mapped to the exact repair procedure step that applies to those readings — not a generic answer, the actual procedure.
The entire system runs on the user's machine. No documentation content, no diagnostic data, and no vehicle information is transmitted to any external server. An internet connection is required for the Claude Desktop AI layer, but all vehicle data and documentation remains on your hardware.
Built and tested on real vehicle documentation. These are capabilities demonstrated in working software, not planned features.
Ask any technical question in plain English across thousands of pages of workshop documentation. Returns the specific procedure, specification, or diagnostic step — with source identified for verification.
Paste a complete data log from any diagnostic platform -- FORScan, Torque Pro, OBD Fusion, HP Tuners, dealer IDS/FDRS exports, AlfaOBD, DashCommand, or any other OBD2 tool -- and receive a variable-by-variable explanation: what each PID measures on that specific engine, whether each reading is normal or abnormal, and how parameters relate to each other.
Given a set of readings, the system identifies exactly where in the diagnostic flowchart those readings place the technician and returns the next procedural step — the same logic a dealership IDS system follows.
Describe symptoms or DTCs and receive the exact parameters to log for your specific scan tool, the conditions to capture them under, and the threshold values that separate normal from abnormal for that specific vehicle. Compatible with FORScan, Torque Pro, OBD Fusion, HP Tuners, dealer-level IDS and FDRS, and any platform that reads OBD2 data.
The documentation library, search index, and all vehicle data remain on the user's hardware. No documentation content is stored on external servers. An internet connection is required for the Claude Desktop AI layer, but the knowledge base itself is fully local.
Tested with 326 documents covering full drivetrain, electrical, fuel, emissions, HVAC, body, and frame systems — 14,601 indexed sections enabling precise retrieval across the complete library.
This technology is built and working. The opportunity is integration with an established documentation platform that already has the subscriber base, the content library, and the distribution infrastructure.
A mutual non-disclosure and non-circumvention agreement is required before demonstration. Licensing and acquisition discussions welcome.
Proof of Concept Results
A live demonstration of the working system is available to qualified parties following execution of a mutual NDA. Use this form to introduce yourself and describe your interest.
Prefer email directly? scott@autokbsystems.com