Final Year B.Tech CSE (AI & Robotics) @ VIT Chennai 💼 Ex-Data Engineer Intern @ TangoEye · 🛰️ Ex-ISRO Intern
💬 Ask me about data engineering, ML/DL, NLP, computer vision, or shipping AI systems to production |  |
Final-year B.Tech (AI & Robotics) engineer specializing in applied machine learning and the production systems that ship it — RAG and LLM-evaluation pipelines, computer-vision models, and the data infrastructure, APIs, and streaming backends behind them. During my Data Engineer internship at TangoEye, I reduced production analytics query latency by up to 84× (65s → 0.77s) with aggregation pushdown and keyset pagination, and built an RBAC-gated operations console managing migration and audit workflows for 6,300+ retail stores on AWS (SQS, Glue, Lambda). My research spans 4 IEEE-indexed publications and 5 patents across IoT, blockchain, and embedded systems, alongside an ISRO internship on launch-vehicle telemetry analysis. I value engineering rigor, measurable performance, and systems that stay reliable under real-world load.
| Role | Core Skills |
|---|---|
| ⚙️ AI Engineer | FastAPI · Docker · AWS (EC2, S3) · CI/CD · MLOps · RAG · LangChain · REST APIs |
| 🤖 Machine Learning | scikit-learn · anomaly detection (Z-Score, Isolation Forest, One-Class SVM) · feature engineering · MLflow |
| 🧠 Deep Learning | TensorFlow · PyTorch · CNNs · Transformers · transfer learning |
| 💬 NLP | RAG · LangChain / LangGraph · Hugging Face · ChromaDB · FAISS · sentence-transformers · RAGAS |
| 👁️ Computer Vision | YOLOv8 · OpenCV · Ultralytics · real-time object detection |
| 🔧 Data Engineering | ETL/ELT · Apache Spark · Apache Kafka · OpenSearch · AWS (SQS, Glue, Lambda) · data warehousing · SQL |
| 📊 Data Science | feature engineering · statistical analysis · A/B testing · pandas · NumPy · Matplotlib · Jupyter |
| 🗄️ Databases | MySQL · PostgreSQL · MongoDB · SQLite · Supabase · schema design · OLAP/OLTP |
| 💻 Software Engineering | Java · C++ · JavaScript · Node.js · Flutter · DSA · OOP · Git · REST APIs |
| Role | Highlights |
|---|---|
| Data Engineer Intern · TangoEye.ai Jun 2026 – Jul 2026 · Chennai, India | Built TangoOps, an access-controlled console managing migration/audit workflows for 6,321 stores across 4 role-gated accounts, replacing direct-AWS shell scripts with atomic, RBAC-gated operations. Reduced production analytics query latency up to 84× (65s → 0.77s) via aggregation pushdown and keyset pagination, decoupled a migration pipeline with SQS, and automated a 5-phase incident fix into one idempotent AWS Glue job with SES/CloudWatch alerting. |
| Project Intern · ISRO – Satish Dhawan Space Centre Jun 2025 – Jul 2025 · Sriharikota, India | Built a Python telemetry app (SQLite3 backend, Tkinter/Matplotlib GUI) for real-time curve visualization and nominal-vs-actual comparison, and integrated an ML anomaly-detection pipeline (Z-Score, Isolation Forest, One-Class SVM) with severity scoring and structured CSV export for post-flight review. |
| Project | Description | Stack | Links |
|---|---|---|---|
| EvalPipe — AI Evaluation Pipeline | Self-hosted, provider-agnostic evaluation pipeline for LLM apps — pluggable metric suite, statistically rigorous A/B testing, and prompt versioning; 95% test coverage, runs fully offline | Python, FastAPI, Docker | Code · Demo |
| Semantic-RAG-Search | RAG pipeline over 18,000+ Usenet posts (1993) — sentence-transformer embeddings in ChromaDB, LangGraph query rewriting + document grading, a Fuzzy C-Means semantic cache (~15× cheaper lookups), and RAGAS evaluation | Python, LangGraph, ChromaDB, FastAPI | Code · Demo |
| TradeWatch | Big-data trade-surveillance engine — Kafka → FastAPI for sub-10ms anomaly alerts (spoofing, wash trades, price/volume spikes) via Z-Score + Isolation Forest, with a Spark + Hadoop (HDFS/MapReduce) layer for distributed detection and backtesting; ~96% recall at <1% false alarms | Python, Kafka, FastAPI, Spark, Hadoop | Code |
| TaskPilot | RL (Q-learning) cloud job scheduler that minimises carbon footprint via tier-pool routing and real-time grid-energy data, with LLM-generated explainability for each scheduling decision | Python, Groq, FastAPI | Code · Demo |
| Curve-Analysis-and-Anomaly-Detection | Desktop tool for launch vehicle telemetry analysis — deviation/bound checks and ML-based anomaly detection. Built at SDSC-SHAR, ISRO | Python, scikit-learn | Code |
| BloodBank-Finder | Real-time blood donation and emergency-response platform — geo-matched donor discovery, live inventory, and ambulance dispatch | Next.js, TypeScript, Prisma, Socket.io, React Native | Code · Demo |
| Coinsight | Cross-platform personal finance tracker — per-category budgets, income/expense logging, savings goals, recurring bills, a financial-health score, and an AI financial advisor (Grok / xAI) that answers purchase/loan/investment decisions from your real data | Flutter, Supabase, Grok (xAI) | Code · Demo |
| Smart-Traffic-System | IoT traffic-signal system that detects approaching emergency vehicles via GPS + siren telemetry and opens a real-time green corridor through intersections ahead of them | React, Node.js, Socket.io, Leaflet, Vite | Code · Demo |
| Research | 4 IEEE-indexed publications across computer vision, predictive analytics, satellite EPS, and maritime safety systems |
| Patents | 5 patents spanning IoT sensing, blockchain-based EHR systems, embedded design, and wearable AR — most recent: Wearable Underwater Navigation & Safety System with Augmented Reality and Multi-Modal Data Fusion |
| Award | 1st Place, 39th National Convention of Electrical Engineers (2024) — Vehicle Distance Estimation using YOLOv8, from 85+ national submissions |