AI / ML Engineer · GenAI & Data Science · Full-Stack AI Systems
I build end-to-end AI systems that combine machine learning, data science, and modern LLM-based workflows.
My work focuses on taking models from data → logic → deployment, with an emphasis on clarity, reproducibility, and real-world usability.
I enjoy working across the stack—from data analysis and predictive modeling to LLM-powered applications and APIs—while keeping systems simple enough to debug, measure, and improve.
- Predictive modeling, forecasting, and experimentation
- LLM applications (RAG, semantic search, personalization)
- Social media & behavioral data analytics
- Full-stack AI systems (API + ML + frontend)
- Clean, production-oriented ML pipelines
A lightweight agent workflow system designed for controlled execution, transparency, and debuggability.
Focus areas:
- Task planning and decomposition
- Short-term and persistent memory
- Structured tool execution
- Frontend-based execution trace visualization
Built to explore reliable agent behavior without opaque abstractions.
A compact, fully local retrieval system designed for semantic search and personalization use cases.
Includes:
- Semantic chunking
- Embedding generation
- Vector indexing with Chroma
- Clean retrieval APIs
Optimized for clarity and extensibility over heavy frameworks.
An applied ML project analyzing Instagram, Twitter/X, and Facebook data.
Key components:
- Sentiment analysis, NER, and theme detection
- Engagement forecasting using time-series models
- Metrics-driven insights for content strategy and scheduling
Demonstrates applied data science, NLP, and business-oriented analytics.
I’m interested in roles that sit at the intersection of AI engineering, machine learning, and data science—especially where models meet real users.