Data Scientist & ML Engineer specializing in large-scale ML pipelines, RAG/LLM applications, and risk analytics on multi-terabyte datasets.
PySpark | Responsible Gambling Modelling | Risk Modelling | RAG/LLMs|
Currently leading risk modeling at Rutgers Center of Gambling Studies, where I work on 4.5 TB / 13.6B+ record datasets to detect high-risk gambling behavior using clustering (KMeans, GMM), gradient boosting (LightGBM), and anomaly detection pipelines.
I also design RAG-powered AI applications for real-time analytics and financial decision-making, delivering systems that scale from prototype to production and directly inform policy and business strategy.
Open to connect for exciting opportunities! 🚀
- 🎓 Master's in Statistics – Data Science at Rutgers University (May 2025)
- 🏢 Currently: Data Scientist at Rutgers Center of Gambling Studies
- ⚡ Previously: Data Analyst at Omalco Extrusion, automated BI pipelines & forecasting systems
- 💬 Ask me about: Responsible gambling analytics, Risk Analytics, Big Data Analytics, RAG systems, time-series forecasting, or real-time dashboards
- 🌱 Exploring: Risk & fraud ML, Agentic AI applications, and RAG systems for domain-specific intelligence
- 📫 Reach me: Email | LinkedIn
- 🌐 Website: Portfolio
📍 Rutgers University – Center of Gambling Studies
Data Scientist | Apr 2024 – Present
Python PySpark K-Means DBSCAN Clustering
- 🚀 Classified 5M+ gamblers with unsupervised ML (Silhouette 0.79)
- 📊 Built ML models (KMeans, GMM, LightGBM, SVM) → 30% boost in early detection
- 🏛️ Results influenced multi-billion dollar state regulations
📍 Omalco Extrusion, India
Data Analyst | Aug 2022 – Jul 2023
Power BI D3.js PostgreSQL ARIMA Prophet
- ⚡ Automated financial reporting → reduced ETL runtime -37%
- 📈 Built sales forecasts with ARIMA/Prophet → improved planning
- 🌐 Designed hybrid PostgreSQL–Couchbase system for live dashboards
For more details, check out my pinned projects! 📌
📚 Full Tech Stack
- Languages: Python, SQL, R, C++, MATLAB
- ML & AI: KMeans, SVM, LightGBM, PyTorch, TensorFlow, HuggingFace, RAG
- Visualization: Power BI, Tableau, D3.js, Matplotlib, ggplot, Plotly
- Cloud & Databases: Google Cloud (BigQuery), Azure, PostgreSQL, Couchbase, Snowflake, Hive
- Other: Outlier Detection, PCA, Time-Series (ARIMA, Prophet), ETL Pipelines
- 🍽️ Fine D-AI-ne → Restaurant QA system with LLaMA3, Mistral, RAG; achieved F1: 0.65, BERT Score: 0.99
- 🐦 Twitter Search Application → Accelerated tweet processing by 60%; hybrid PostgreSQL–Couchbase–Elasticsearch architecture
- 🗺️ Campus Watch → Real-time geospatial dashboard using Selenium + Plotly for campus crime analytics