- 🎓 MSc Computer Science graduate (CGPA 7.20) — Dr. Babasaheb Ambedkar Marathwada University
- 🏗️ Built 31 production-grade, end-to-end ML systems — not notebooks, but full MLOps pipelines
- 🔧 Every project ships with: FastAPI serving, Streamlit monitoring dashboards, Champion-Challenger model governance (3-gate promotion), SHAP explainability, PSI drift monitoring, MLflow tracking, Docker + GitHub Actions CI/CD
- 🌐 Domains covered: BFSI, healthcare, retail, computer vision, NLP, recommender systems, GenAI / RAG
- 💼 Open to Data Scientist / ML Engineer roles
| Project | Domain | Model | Key Metric |
|---|---|---|---|
| DriveGuard AI | Computer Vision / Smart Cities | YOLOv8m + DeepSORT + Gemini RAG | mAP@0.5 = 0.680 · 364 tests |
| AI Loan Underwriting Platform | BFSI | XGBoost (credit + fraud) | ROC-AUC 0.858 / 0.955 · 242 tests |
| CineMatch Pro | Media Tech | SVD + TF-IDF + SBERT · CLIP · RoBERTa | RMSE 0.8005 · Acc 94.26% |
| MathVision AI | EdTech / CV | TrOCR-large + SymPy | CER 0.138 · 350/350 solved |
| AgriSense AI | AgriTech | EfficientNetB3 | Acc 98.44% · 38 classes |
| HR Policy Chatbot | GenAI / RAG | Gemini + ChromaDB | 30/30 tests |
| BC5CDR Clinical NER | Biomedical NLP | BioBERT | F1 = 0.8847 |
| Credit Card Fraud Detection | BFSI / Fintech | ExtraTrees | F1 0.8962 · PR-AUC 0.8817 · 284K transactions |
📌 All 31 projects: see the repositories tab — Credit Risk, Churn, House Price, Sales / Energy / Stock forecasting, Resume Screener, ABSA, Fake News, Topic Modeling, Chest X-Ray, Object Detection, Emotion Recognition, Customer Segmentation, Market Basket, Recommender, Readmission, Attrition, MNIST-from-scratch, Insurance, Cross-Sell, Ride Fare, BI / SQL.
🌐 Portfolio: narendrakalam2001.github.io