Building intelligent systems at the edge of data & possibility
I'm a Data Scientist & ML Engineer with 5 years of experience and a Master's in Computer Information Systems from Boston University (GPA 3.85). My work spans classical ML, large language models, agentic AI, and production-grade generative AI systems.
At Route Mobile I owned the full ML lifecycle — from building and monitoring Airflow-powered data pipelines to shipping customer segmentation (KMeans) and demand forecasting (XGBoost) models. I love working at the intersection of LLMs, RAG architectures, and autonomous agents.
When I'm not training models, I'm designing systems that think, reason, and act.
Real-time intruder detection system using YOLOv8 object detection combined with MOG2 background subtraction. Processes live video streams with OpenCV, filtering false positives and triggering intelligent alerts.
Autonomous trading agent using LangGraph for stateful decision-making. Integrates financial data tools, reasoning chains, and observability via LangSmith for end-to-end monitoring of agent behavior.
Production-grade retrieval-augmented generation chatbot. Combines FAISS vector search with GPT-4o for grounded, context-aware responses. Built with a clean Streamlit interface for real-time interaction.
Collaborative filtering recommendation engine deployed on AWS EMR. Leverages PySpark's ALS algorithm for matrix factorization across millions of user-item interactions with distributed compute.
Fine-tuned BERT-based NLP classifier for intent detection and content categorization. Applied at Route Mobile for real-world telecom messaging use cases with high-volume production traffic.