* Built AI-powered learning tools (flashcards, AI document editor, STEM video explanations).

* Implemented an LLM fallback architecture to improve reliability and reduce API costs.

* Developed engagement systems (streaks, milestones) and token-to-credit usage model.

* Built internal tooling (Vercel log viewer, GitHub Actions deploy workflow) to reduce platform seat costs.

- Increased Sonar test coverage by 60% across a core repository through unit and integration testing.

- Resolved critical Snyk and Sonar vulnerabilities to improve reliability and security posture.

- Worked on event correlation tooling to improve observability and debugging of distributed systems.

- Built a medical search platform frontend using Next.js and TypeScript to improve usability and performance.

- Migrated architecture from React (Vite) to Next.js to improve performance and reduce build complexity.

- Integrated an AI medical chatbot with an ICD-10 dataset for faster diagnosis lookup and information retrieval.