Kaviya Rathinakumar

Emerging AI Engineer and Full-Stack Developer with

1+ year of hands-on experience building production-grade

RAG (Retrieval-Augmented Generation) pipelines,

LLM-powered applications, and scalable web platforms.

Skilled in Python, FastAPI, React.js, vector database integration, semantic search, and prompt engineering.

Proven ability to deliver end-to-end solutions across freelance and internship engagements — from backend API design to AI-integrated document retrieval systems.

AI Engineer Intern at Quibo Tech (2025-12 – Present)

ArchiReach — Intelligent Document Retrieval Platform

• RAG Pipeline Development: Engineered end-to-end Retrieval-Augmented Generation pipelines with LangChain-style retrieval, improving response relevance by 30%+ through optimised vector search, semantic chunking, and embedding model tuning
• Python Backend Services: Built and maintained Python-based document ingestion, preprocessing, and vector store indexing services, ensuring low-latency query responses at production scale
• Prompt Engineering & AI Architecture: Designed modular AI solution architectures applying prompt engineering best practices to reduce hallucinations, improve answer grounding, and deliver robust LLM-powered outputs
• Performance Optimisation: Integrated optimised data retrieval mechanisms and embedding pipelines, measurably improving system performance and end-user experience

Freelance Full-Stack & AI Developer at Neoyond (2025-04 – 2025-11)

Full-Stack Web & AI-Integrated Applications

• Full-Stack Development: Delivered end-to-end web applications using Python (FastAPI) on the backend and React.js on the frontend, owning the full development lifecycle from scoping through deployment
• RAG-Powered Features: Built document Q&A systems using semantic search, vector embeddings, and LLM-generated responses, enabling clients to query private knowledge bases and eliminating manual lookup workflows
• Database & API Design: Designed PostgreSQL and MongoDB schemas with indexing, query optimisation, and Pydantic validation; developed and documented RESTful APIs consumed by React.js frontends
• React.js UI Development: Built reusable component libraries with hooks-based state management, Axios API integration, and responsive Tailwind CSS layouts for cross-device compatibility

B.Sc. in Computer Science – The Gandhigram Rural Institute (2022 – 2025)