Engineered core platform functionality for an internal online learning project, using TypeScript, Next.js, Supabase (PostgreSQL), Sanity (NoSQL), and a custom Retrieval-Augmented Generation pipeline; owned architecture decisions, schema design, and reliable session/state management across the stack.

Built a fully modular REST API layer from the ground up, standardizing routing, error handling, authentication, and data-access patterns; reduced integration complexity and improved system reliability and observability.

Led development from requirements → technical design → production deployment, operating with high autonomy while aligning closely with stakeholders on UX, data integrity, and long-term scalability.

Strengthened performance and maintainability by introducing clean architecture patterns, normalized data flows, and predictable state transitions, setting the foundation for future ML-powered personalization features.

Building internal platforms and AI-driven systems at PromptLayer to streamline workflows and enhance developer productivity. Working across backend and frontend to design, implement, and scale tools that support rapid experimentation and deployment of LLM-powered applications.

Worked on a wide range of in-house initiatives both independently and in close collaboration with cross-functional teams. Took full ownership of end-to-end feature development, from scoping and architectural design through implementation, testing, and deployment. Built and maintained full-stack applications using the Next.js framework, integrating TypeScript, React Server Components, API routes, and optimized data-fetching patterns. Contributed to system reliability and scalability by improving internal tooling, refactoring shared components, and designing modular infrastructures that supported rapid iteration across the team.

Served as the lead front end engineer, owning the architecture and development of major components of the company web platform. Took responsibility for defining requirements, conducting technical research, and delivering production ready features with minimal oversight. Built performant and maintainable interfaces using Vue, TypeScript and JavaScript, Node, Quasar, and Apache E Charts while improving overall usability and reliability across the platform.