Manufacturing Supply Chain AI App

I run a manufacturing operation that needs a web-based application powered by AI to keep our supply chain running at peak efficiency. The core objective is supply chain optimization—specifically, I want the system to learn from our historical and real-time data, flag bottlenecks before they appear, and recommend concrete actions my planners can take.

Data the models must digest and correlate:
• Inventory levels
• Supplier performance metrics (on-time delivery, quality scores, lead-time variability)
• Demand-forecasting inputs from sales and market signals

Key workflow I have in mind:
1. Secure data ingestion from our existing ERP/MES via API or flat-file drops.
2. Automated preprocessing, feature engineering, and model training (Python, TensorFlow/PyTorch, or an equivalent stack).
3. Interactive dashboards in a responsive web interface—React, Vue, or a similarly modern front-end—showing predictive KPIs, alerts, and “what-if” scenario tools.
4. Role-based access so planners, buyers, and execs each see the insights that matter to them.
5. Clear documentation plus a hand-off session so my internal IT team can maintain and retrain the models.

Acceptance criteria:
• Forecast accuracy and supplier risk scores must outperform our current baseline by at least 15 %.
• Page loads and model inferences return in under two seconds on typical data volumes.
• Codebase pushed to a private Git repo with clean commits and inline comments.

If you’ve delivered similar AI supply-chain solutions before and can take this from architecture through deployment on our preferred cloud (AWS or Azure), let’s jump in—I’m ready to provide sample datasets and sandbox credentials right away.