ML Engineer – Predictive Insights for Amazon Orders & Shipping

Job Description
• We’re an Amazon-only store with years of order + full shipping-journey data (scan-by-scan, customer chats, returns, A-to-Z outcomes, pricing, margins… everything).

• Your mission: turn that raw data into “see-the-future” insights, e.g.
• predict A-to-Z claim risk, late deliveries, returns/refunds, customer-first messages
• rank SKUs by real ROI (after refunds, shipping risk, claims)
• surface supplier / carrier reliability hot-spots
• flag one-off anomalies (fraud, bad addresses, repeat abusers)

What you’ll actually do
• wrangle + feature-engineer our dataset (SQL / Python / Pandas)
• train & validate ML models (LightGBM / XGBoost or your weapon of choice)
• expose order-level risk scores through a lightweight API or file drop
• hand us simple dashboards or Google Sheet views for the ops team
• document everything so we can retrain / tweak without you next time

Why this is friendly for you:
• We know our data cold and can explain business logic fast—no industry ramp-up needed.
• Clean schema access, labelled outcomes, and a founder who answers questions in minutes.

Must-have skills
• Python ML stack, SQL, basic DevOps (Docker or serverless deploy)
• Proven track record of shipping models into production, not just notebooks

Nice-to-have
• Experience with e-commerce, Amazon SP-API, or logistics datasets
• How to apply
• Send CV / GitHub + 2-3 sentences on a similar project you owned
• Bonus points for a link to a live demo or repo with an API endpoint