I'm an AI Engineer & Backend Engineer at Tata Consultancy Services, where I design agentic RAG systems and production-grade microservices that process real enterprise workloads — think 100K+ invoices a month, 50,000+ policy records analyzed for drift in real time, and LLM-powered agents that reason, retrieve, and act with human-in-the-loop oversight.
I sit at the intersection of Generative AI and distributed backend systems — building the LangChain/LangGraph agent on one side and the Kafka-driven, Kubernetes-deployed infrastructure that keeps it reliable at scale on the other.
- 🔭 Currently building: Agentic RAG for health insurance, AI-powered ITSM triage & recommendation systems, and event-driven microservices modernizing claims/invoice processing
- 🌱 Currently deepening: Fine-tuning (LoRA/QLoRA), multi-agent orchestration, and inference optimization
- 🎯 Known for: Turning "top 5 out of 400+ teams" hackathon prototypes into production-viable systems that get COO-level recognition
- 🤝 Open to: Mentoring, technical talks on GenAI/Agentic AI, and backend architecture collaborations
|
🥇 TCS AI Hackathon — Top 5 / 400+ teams Architected an LLM-powered compliance tracker generating SQL rules from natural-language policy docs via RAG, with human-in-the-loop approval. 🎖️ TCS AI Friday — Regional Finals, Top 5 Innovation challenge win with COO executive recognition; solution showcased across multiple customer accounts and the TCS AI org. |
📢 Trained 400+ TCS associates Delivered enterprise-wide sessions on Generative AI, prompt engineering, fine-tuning, and AI-driven development practices. 👨🏫 Mentored 200+ students Led Python, OOP, and AI/ML workshops at SNS College of Engineering on system design and ML fundamentals. |
| Metric | Impact |
|---|---|
| ⚡ Invoice processing throughput | 3x improvement via event-driven redesign |
| 🔁 Regression test cycle time | 4 days → 6 hours (95% reduction) |
| 🐢 API response latency | 1.2s → 180ms via query/index tuning |
| 🚨 Mean time to detection (MTTD) | 30 min → under 2 min |
| 🔄 Data sync latency (CDC pipeline) | ↓ 60%, 10,000+ records, zero downtime |
| 🎯 ITSM operational efficiency | ↑ 40% via automated data quality engine |
| 🛡️ Deployment reliability | 99.9% SLA with blue-green Kubernetes releases |
| 📄 Documents processed monthly | 100,000+ across AI pipelines |
Languages
Generative AI & ML
Backend & Architecture
Data & Vector Stores
Cloud & DevOps
Observability
|
🧩 Multimodal Computer Interface Adaptive HCI system fusing Gemini Vision, Whisper, and text for cross-modal task execution — 95% task completion accuracy, sub-200ms response times.
|
⚙️ Adaptive AI Data Quality System Auto-generates SQL validation rules from SOPs; anomaly detection across 50,000+ records with Isolation Forest + vector search — 80% less manual review.
|
📋 LLM Compliance Tracker Natural-language policy → SQL rule generation with RAG retrieval and human-in-the-loop approval. Top 5 of 400+ at TCS AI Hackathon.
|
|
🤖 Human-in-the-Loop Agent Interface Local review UI serving AI-generated plans/artifacts with modular components for diagrams, tables, code, and forms — plus a portable HTML export pipeline.
|
📨 AI Resume Tailoring Platform Event-driven job sourcing pipeline (Kafka + Spring Boot + Python scraper) paired with a Gemini-powered engine that dynamically tailors resumes to JDs.
|
🎫 AI-Powered ITSM Solution Intelligent ticket triage via NLP classification, semantic-search recommendations, and a data quality engine — 40% efficiency gain.
|
B.Tech, Artificial Intelligence & Data Science — SNS College of Engineering, Coimbatore (May 2024, CGPA 8.5/10) Coursework: Deep Learning · NLP · Computer Vision · Reinforcement Learning · Distributed Systems · System Design
I'm open to mentoring, speaking on GenAI/Agentic AI and backend architecture, and collaborations on production-grade AI systems.
Topics I can speak on: Agentic RAG design · Prompt engineering & fine-tuning · Event-driven microservices · MLOps & observability · Scaling LLM applications in production
⭐️ From ideation to production — building AI systems that ship. Last updated: July 13 2026