CAUSYN AI — Predict. Understand. Prevent.

**CAUSYN AI — Predict. Understand. Prevent.** Unexpected industrial machine failures can interrupt production, increase maintenance costs, and leave engineers making difficult decisions with limited information. CAUSYN AI addresses this challenge by bringing machine-risk prediction, AI-assisted investigation, hypothetical simulation, and human maintenance decisions into one platform. **Predict:** CAUSYN AI uses a Random Forest machine-learning model trained on the AI4I 2020 synthetic predictive maintenance dataset. Engineers can scan a fleet of machines, inspect sensor readings, and identify assets with elevated model risk scores. The current prototype includes eight simulated machines and supports registering additional machines. **Understand:** CAUSYN AI combines sensor-based evidence and rule-based hypotheses with generative AI powered by AMD Token Factory. It produces structured investigation reports and offers an interactive AI Copilot that helps engineers decide what to inspect. Suggested contributing factors are hypotheses, not proven root causes. **Prevent:** Engineers can use What-If Simulation to explore how hypothetical sensor changes affect the model's risk score. A Maintenance Decision Center also supports reviewing, approving, deferring, or rejecting proposed actions, keeping engineers in control. **Technology:** Next.js, React, Python, FastAPI, scikit-learn, Random Forest, AMD Token Factory, Docker, and Render. **Current scope and future vision:** CAUSYN AI is a deployed, working hackathon prototype demonstrated with simulated industrial data. Its risk scores are uncalibrated model estimates, not verified failure probabilities. Future development would integrate real factory sensor data, validate models in industrial environments, and customize the platform for different companies and their specific machines.