CASE STUDY

AutoCraft

AI-powered mechanical routing system & diagnostics visualizer.

Project Overview

AutoCraft was built to bridge the gap between automotive telemetry and standard vehicle operators. The system uses a trained intent classifier to map engine diagnostics and code readouts into actionable mechanical repairs, allowing drivers to diagnose issues immediately.

System Architecture

[ OBD-II Telemetry Data ] │ ▼ (POST /api/diagnose) [ FastAPI Backend ] │ ├─► [ TensorFlow NLP Intent Classifier ] │ │ │ ▼ │ (Extract Severity & Fault Coordinates) │ └─► [ Mechanics Routing Matrix ] │ ▼ [ Diagnostic Visualizer Client ]

The frontend is crafted with React and Tailwind to draw interactive vector parts overlays, while the backend runs FastAPI pipelines routing diagnostic arrays to classification trees in sub-120ms.

Challenges & Breakthroughs

Dynamic Vector Overlay Coordinates

Mapping irregular technical coordinates on mechanical schematic overlays dynamically on viewport changes. Solved by building responsive SVG coordinate calculations based on fluid container dimensions.

Technologies Used

ReactPythonFastAPITensorFlowTailwind CSS

Project Specs

TypeAI Mechanical Engine
DatabaseRelational DB
DeveloperSaumyadeep C.