Sequoia Applied Technologies
Chennai
Built a semantic search RAG platform(FastAPI + React) using Azure AI Search vector embeddings for retrieval across Azure SQL, NoSQL, and JSON sources, enabling natural language querying and cutting cross-database query time by 60%.
Shipped a production MCP server for GenXFlo with API key authentication and MCP Resources, enabling AI agents to autonomously trigger and orchestrate enterprise workflows end-to-end — eliminating manual intervention across client processes.
Architected an AI-Driven SDLC Framework integrating GitHub Copilot, Jira, Confluence, and CI/CD pipelines to automate the full delivery lifecycle — requirement gathering, code generation, testing, and documentation — for client projects.
Sequoia Applied Technologies
Chennai
Built an LLM-powered automated test case generation pipeline (OpenAI APIs + Jira) for POS systems across diverse hardware platforms, reducing manual testing effort by 70% and accelerating client delivery.
Engineered AI-driven interview question generation in a recruitment platform using advanced prompt engineering, improving question relevance by 60% for enterprise clients.
Enkefalos Technologies LLP
Mysore
Shipped an employee activity tracking system (Django, REST API, Selenium, PostgreSQL) with a 30% improvement in real-time productivity monitoring for enterprise clients.
Built RESTful APIs for employee data and attendance, cutting manual data management by 40% across 60+ users.
Analysed large-scale activity datasets to surface productivity bottlenecks, delivering insights that drove a 20% increase in workplace efficiency.
B.Tech
Specialization: AI & Machine Learning
CGPA: 8.66
Senior Secondary Education
81%
Secondary Education
82%
Built Travel Optix— a multi-agent system using CrewAI, browserless, and Serper APIs to generate personalised itineraries in under 5 minutes, outperforming leading LLM-based planning tools by integrating real-time weather, cost, and event data.
Combined YOLOv7–YOLOv9 with Deep SORT for real-time traffic monitoring; achieved 90.4% mAP across helmet detection, speed estimation, and number plate recognition.
Built an end-to-end multi-agent system using CrewAI, LangChain, and Serper APIs to autonomously generate personalised travel itineraries in under 5 minutes, integrating live weather, cost, and event data.
Designed agent orchestration and real-time data pipelines that outperformed GPT-4 and Gemini on itinerary quality benchmarks; published as a research paper.
Shipped a production-grade RESTful API with JWT auth and real-time tracking via Binance WebSocket, achieving 96% alert accuracy for 100+ active users.
Deployed with Redis caching and Docker for operational stability at scale, reducing notification latency by 20%.