Applied AI and machine learning engineer. I build end-to-end systems: data pipelines, NLP, retrieval-augmented generation, and the interfaces on top.
Selected projects
- osint-ai: ingests public news RSS, extracts entities with spaCy, maps how they co-occur as a graph, and answers questions over the corpus with local RAG. Runs fully local, no API keys.
- warranty-prediction-agent: AI agent for warranty prediction built on Google ADK, Gemini, and BigQuery ML.
- customer-churn-sentiment-analysis: end-to-end ML and NLP, churn prediction with Random Forest and SMOTE plus social-media sentiment analysis.
- dora-metrics-dashboard: native macOS dashboard for DORA metrics and PR review times, built with SwiftUI.
- where: full-stack location-based social app, React frontend with a Java Spring backend.
Working with Python, RAG, LangGraph, spaCy, ChromaDB, sentence-transformers, Streamlit, Docker, GCP, Terraform.
I care about responsible AI and honest evaluation of what models can and cannot do.
