Naveenkumar N
Computer Science Engineer building full-stack systems and AI applications.
I build practical software across Full-Stack Development • AI • Backend Systems
What I actually work on
I'm a Computer Science student at Madras Institute of Technology, focused on building full-stack applications, AI-powered systems, and reliable backend services.
My work spans everything from authentication-driven web applications to biomedical AI systems using knowledge graphs and LLMs. During my internship at AstraZeneca, I worked on a biomedical AI assistant combining knowledge-graph retrieval with literature-based evidence.
I enjoy working on problems where software needs to be accurate, scalable, and reliable—not just functional.
Where I've worked
AstraZeneca
- Built a biomedical AI assistant using React, FastAPI, Neo4j, and Pydantic AI for knowledge-grounded retrieval.
- Developed LLM workflows that generate and validate Cypher queries against a biomedical knowledge graph.
- Combined structured knowledge-graph results with biomedical literature to produce evidence-backed responses.
- Designed automated evaluation workflows measuring accuracy, concept coverage, similarity, and response quality.
Projects, as spec sheets
Pet Adoption System
- User and admin authentication with role-based access
- Category-based search and adoption history tracking
- Admin dashboard for managing listings and requests
Canteen Menu Management API
- Secure login, password hashing, token validation, and role-based access via Spring Security.
- Spring Data JPA with MySQL, DTO validation, and centralized exception handling.
- Clean service-layer architecture for a reliable, maintainable backend.
Post-Quantum Secure Authentication for 5G N1 Interface
- ML-KEM for key exchange and ML-DSA for digital signatures, replacing classical primitives vulnerable to quantum attacks.
- Evaluated latency, bandwidth usage, and computational cost against baseline authentication.
- Assessed feasibility for future 5G deployments under real network constraints.