About Me
_ The Engineer
I approach software engineering with a systems-first mindset. My expertise lies in architecting robust backends, designing clean RESTful APIs, and building secure authentication flows. I focus on writing maintainable, tested code that solves real scalability challenges.
I specialize in integrating AI into production environments. From designing efficient RAG pipelines to managing vector embeddings and orchestrating agentic workflows, I treat AI models as reliable software components rather than black boxes.
While I prioritize backend logic, I am fully proficient in the modern frontend stack. I use tools like React and Next.js to build performant, type-safe interfaces that serve as the delivery mechanism for complex underlying systems.
Selected Work
Engineering scalable systems and solving complex problems with code.
Mini Agentic Reasoning Framework
A modular framework for building autonomous agents capable of planning, reasoning, and tool execution. Designed to handle ambiguous instructions and dynamic state management.
◆ Backend Highlights
- Implemented Hierarchical Task Network (HTN) for decomposed planning
- Built a robust state machine for managing agent context and memory
- Designed a plugin system for dynamic tool loading and execution
eVALuate
An automated evaluation platform that uses LLMs to grade student submissions, providing detailed feedback and competitive analysis.
◆ Backend Highlights
- Designed scalable microservices architecture for concurrent grading jobs
- Implemented RAG pipeline for rubric-based consistency checking
- Optimized prompt engineering to reduce grading costs by 40%
Nexis
A privacy-first AI companion that processes text and voice to provide real-time emotional support and resource recommendations.
◆ Backend Highlights
- Developed secure websocket pipelines for real-time voice processing
- Implemented local PII redaction before data leaves the client
- Designed an event-driven architecture for scalability
LinkedIn Talent Finder
A high-performance search tool that ingests resumes and ranks them based on semantic similarity to job descriptions using vector search.
◆ Backend Highlights
- Built an ingestion pipeline parsing PDF/DOCX at 50+ files/sec
- Implemented hybrid search (Keyword + Vector) for higher relevance
- Optimized Qdrant vector database queries for sub-100ms latency
Get in Touch
I am currently open to new engineering roles and interesting conversations. Building high-impact systems is what I do best.