I built LLM Cost Tracker & Advisor, a full-stack platform that calculates API costs across OpenAI, Anthropic, and Groq without admin-level API access. It also provides usage-pattern advice through a REST API and MCP.
At Bizom, I engineered six REST API integration services across four enterprise clients and wrote SQL validation logic to catch data discrepancies. I also built and evaluated deep learning models for dysarthric speech classification and reconstruction, reaching 96% classification accuracy.

