I’m an AI engineer and M.S. student at SJSU, focused on building LLM-based agents and the systems around them. I also enjoy working across the ML pipeline from cleaning and preparing data to training, evaluating, and deploying models. Always happy to talk shop. Let’s talk.
Selected Projects
ML pipeline for bankruptcy risk — benchmarked Logistic Regression, Random Forest, and XGBoost across 9 class-imbalance techniques (F1 +40–60%), with threshold/cost analysis and isotonic-calibrated risk scores.
Cloud analytics pipeline on GCP — a Streamlit uploader lands files in Cloud Storage, a Cloud Function loads BigQuery via a star schema (40% faster queries), surfaced in a 7-chart Looker Studio dashboard.
Full-stack social film catalogue — ratings, reviews, diaries, ranked lists, feeds, and recommendations across 25+ REST endpoints. Colocating backend and DB in AWS us-east-1 cut query latency from ~70 ms to ~1 ms.
GPU-free RAG pipeline for research-paper Q&A and summarization — semantic retrieval over FAISS grounds Llama 3.1 (via Groq) with source citations and map-reduce summaries for long documents.
Stack
Contact
Open to AI engineering roles and collaboration. Reach out about technical work or interesting problems.