Mohammad Ahmadian

Summary

AI / Machine Learning Engineer focused on production-oriented ML and applied AI systems. Flagship public portfolio of five Released systems with evaluation metrics, FastAPI serving, Docker packaging, and hiring presentation packs.

Core skills

• Machine Learning: Scikit-learn, LightGBM, evaluation discipline, leakage control
• Deep Learning / CV: PyTorch, classification, detection, segmentation, Grad-CAM
• NLP / LLMs: Embeddings, RAG patterns, NL→SQL, citation-oriented systems
• MLOps / Backend: FastAPI, Docker, MLflow, pytest, GitHub Actions

Flagship projects

• Production ML Platform for Predictive Maintenance — Released

  End-to-end failure-risk scoring: leakage-safe features, chronological evaluation, LightGBM vs baseline, FastAPI serving.

• Enterprise Document Intelligence Platform — Released

  IDP with ingest, retrieval, RAG citations, field extraction, evaluation harness, and FastAPI delivery.

• Enterprise Business Analytics & Forecasting Platform — Released

  Chronological KPI forecasting with baselines vs LightGBM, MAPE/RMSE reporting, and FastAPI decision support.

• AI SQL Copilot — Released

  Schema-aware NL→SQL with validation, explanation, and safe SELECT-only execution — not a chatbot wrapper.

• Computer Vision Inspection Platform — Released

  Multi-task industrial inspection: classification, segmentation, segment-then-box detection, Grad-CAM, FastAPI.