I build AI systems that actually work in production, not just in Jupyter notebooks. Over the past 5 years, I've gone from training demand forecasting models at Unilever to designing full-stack agentic AI applications powered by large language models. My work sits at the intersection of machine learning engineering, LLM orchestration, and real-world deployment and I care deeply about the gap between "demo" and "done." At Unilever, I built ML models in PyTorch and TensorFlow that forecasted demand across 20+ product SKUs, cutting inventory overstock by 18% and improving prediction accuracy by 15% through systematic feature engineering and hyperparameter tuning. I also deployed those models to production via REST APIs and Docker on AWS, since a model that never ships doesn't help anyone. At Governors State University, I led data pipeline development supporting predictive analytics across 100,000+ institutional records, automated reporting dashboards that saved 30% of manual effort, and improved operational planning accuracy by 22% through advanced statistical forecasting. On the project side, I've built: → An AI Agent Copilot using LangChain, FastAPI, and RAG pipelines with FAISS/ChromaDB enabling multi-turn reasoning, conversation memory, and document-grounded responses. → MediGPT, a healthcare LLM assistant with HIPAA-aligned guardrails, semantic retrieval, and real-time low-latency response via FastAPI. → A RAG Document Search Engine on AWS Bedrock + LangChain letting non-technical users query unstructured documents in natural language. My stack: Python · PyTorch · TensorFlow · LangChain · RAG · FastAPI · Docker · AWS (EC2, S3, Bedrock) · FAISS · ChromaDB · SQL · Apache Spark · Power BI I'm currently finishing my MS in Computer Science at Governors State University (GPA: 3.9/4.0) and am actively looking for AI Engineer, ML Engineer, or Applied Scientist roles where I can keep building things that matter. If you're working on AI-powered products and want to talk shop or if you have a role where LLMs, RAG, or agentic systems are central - let's connect.
I build AI systems that actually work in production, not just in Jupyter notebooks. Over the past 5 years, I've gone from training demand forecasting models at Unilever to designing full-stack agentic AI applications powered by large language models. My work sits at the intersection of machine learning engineering, LLM orchestration, and real-world deployment and I care deeply about the gap between "demo" and "done." At Unilever, I built ML models in PyTorch and TensorFlow that forecasted demand across 20+ product SKUs, cutting inventory overstock by 18% and improving prediction accuracy by 15% through systematic feature engineering and hyperparameter tuning. I also deployed those models to production via REST APIs and Docker on AWS, since a model that never ships doesn't help anyone. At Governors State University, I led data pipeline development supporting predictive analytics across 100,000+ institutional records, automated reporting dashboards that saved 30% of manual effort, and improved operational planning accuracy by 22% through advanced statistical forecasting. On the project side, I've built: → An AI Agent Copilot using LangChain, FastAPI, and RAG pipelines with FAISS/ChromaDB enabling multi-turn reasoning, conversation memory, and document-grounded responses. → MediGPT, a healthcare LLM assistant with HIPAA-aligned guardrails, semantic retrieval, and real-time low-latency response via FastAPI. → A RAG Document Search Engine on AWS Bedrock + LangChain letting non-technical users query unstructured documents in natural language. My stack: Python · PyTorch · TensorFlow · LangChain · RAG · FastAPI · Docker · AWS (EC2, S3, Bedrock) · FAISS · ChromaDB · SQL · Apache Spark · Power BI I'm currently finishing my MS in Computer Science at Governors State University (GPA: 3.9/4.0) and am actively looking for AI Engineer, ML Engineer, or Applied Scientist roles where I can keep building things that matter. If you're working on AI-powered products and want to talk shop or if you have a role where LLMs, RAG, or agentic systems are central - let's connect.