Data Scientist | 4+ YOE | Production ML & Analytics
Cut Mercedes QA time 60% • Improved forecasting 28% • Built GenAI compliance workflow
ML/AI: Python • XGBoost • PyTorch • LangChain • FAISS • ARIMA/LSTM
Data: SQL • PySpark • Airflow • Databricks • MLflow
Cloud: AWS • Azure • GCP
| Project | What It Does | Impact | Tech |
|---|---|---|---|
| OTT Churn Copilot | ML + GenAI for customer retention | +6.1pp retention lift | LightGBM • LangChain • Streamlit |
| Lang2Query | Text-to-SQL translation | 77% ROUGE improvement | Mistral-7B • LoRA • RAG |
| Last Mile Optimizer | Delivery network simulation + ML dispatch | Improved SLA compliance | XGBoost • Discrete Event Sim |
| Conversation Analytics | Governed warehouse + BI layer | Decision-ready insights | SQL • LLM enrichment |
| Ads ML POC | Retrieval + Ranking + Brand Safety | End-to-end serving demo | Two-tower • CTR model |
Open to Data Scientist and Decision Analytics roles
📧 oberoiharshith8@gmail.com
💼 LinkedIn
📍 New Jersey, USA
Building ML systems that ship 🚀