AI & Data Engineer focused on building LLM systems that ship to production — from RLHF pipelines and multi-agent architectures to GCP data infrastructure.
Currently, shipping RLHF-grade training data and agentic LLM systems for Microsoft, Moonshot AI (KIMI), and Penguin AI.
LLM Systems & Agents Adversarially evaluated GPT-o1, GPT-4o, Gemini, Claude, and DeepSeek across 6 domains. Designed multi-turn agentic workflows (LangGraph, LangChain) with sequential/parallel tool calls. Built self-correcting RAG pipelines with 40%+ performance over single-agent baselines.
Data Engineering Migrating 4+ legacy pipelines to GCP (Kafka, Airflow, BigQuery, medallion architecture). Building an AI cybersecurity log analyzer for plain-English anomaly detection over enterprise data warehouses.
AI Product Consulting Consulted 10+ clients across fintech, SaaS, and analytics at Ember AI — end-to-end from discovery to shipped product. Key delivery: NL data query chatbot over enterprise warehouses built with LangChain and RAG.
| Layer | Tools |
|---|---|
| LLMs & Agents | LangChain, LangGraph, RAG, RLHF/SFT, tool-call design |
| Data Engineering | Kafka, Airflow, BigQuery, GCP, ETL, medallion architecture |
| ML/AI | PyTorch, TensorFlow, Scikit-learn, FastAPI, OpenCV |
| Core Stack | Python, SQL, React, Node.js, Docker |
Open to remote full-time roles and freelance/contract engagements in LLM systems, AI engineering, and data infrastructure.