GenAI Engineer | M.Tech CS @ NIT Durgapur | Ex-HCLTech SDE

I'm a GenAI Engineer specializing in Large Language Models and RAG architectures, currently pursuing my M.Tech in Computer Science at NIT Durgapur (CGPA: 9.0). With 2+ years of industry experience at HCLTech and recent work with the Indian Army on AI-driven predictive systems, I focus on building production-grade AI applications that solve real-world problems.

My work centers on developing end-to-end LLM solutions using modern frameworks and deploying scalable systems that deliver measurable impact.

May 2025 – July 2025 | Panagarh Military Station

• Designed and deployed an AI-powered predictive maintenance system for 1,000+ military vehicles, improving maintenance efficiency by 65%
• Built real-time analytics platform using Streamlit-FastAPI microservices with ML pipelines for clustering, regression, and anomaly detection
• Engineered ETL workflows for time-series forecasting and vehicle health scoring, enabling proactive maintenance decisions

July 2022 – November 2023 | Noida

• Developed enterprise-grade microservices using Java, Spring Boot, and Hibernate for financial systems serving millions of transactions
• Led development of Bank of India's Workitem Approval System with multi-level workflows and Kafka-based asynchronous processing
• Reduced end-to-end latency by 30% through optimized service discovery, API design, and load-balanced routing
• Improved system reliability through SQL optimization, caching strategies, and CI/CD automation

Tech Stack: FastAPI, SQLAlchemy, MySQL, OSRM, React 19, TypeScript, deck.gl, MapLibre GL, Recharts, Tailwind CSS

Built a full-stack GPS distance-intelligence platform that ingests raw telematics data, segments trips, and produces five independent distance measurements per journey — including chunked parallel OSRM map matching (95-point windows, 8 concurrent workers) with synthetic haversine fallback. The system spans 74 API endpoints, 21 database tables, and a 9-page React SPA featuring a real-time Command Center with deck.gl WebGL rendering, SSE-driven vehicle tracking, on-map trip playback (1×–100× speed), and a 24-hour scrubbable timeline. Implemented pure-Python O(n³) Hungarian assignment, 2-opt TSP refinement, DBSCAN clustering, DTW/Fréchet route similarity, 5-component driver scoring, corridor adherence analysis, and a public Distance API (v1) processing up to 20K GPS pings per call. Backed by 100 automated tests validating every algorithm against scipy/scikit-learn reference implementations.

Tech Stack: FastAPI, XGBoost, LightGBM, Isolation Forest, scikit-learn, MySQL, React, TypeScript

Engineered a tiered ML-as-a-Service platform serving 10 production models — ETA prediction, SLA classification, anomaly detection, driver fatigue monitoring, GNN-based route optimization, and demand forecasting — trained on 7.4M+ trip records across 780K drivers and 342K vehicles. Designed a subscription API with 3-tier authentication (Basic / Pro / Enterprise), automated daily/weekly/monthly retraining pipelines, and a React dashboard with 15+ interactive pages. Achieved sub-200ms inference latency with 22-factor feature engineering per prediction.

Tech Stack: LSTM, GRU, CNN, Python, MongoDB, Streamlit

Built an AI-driven predictive maintenance platform analyzing 10 years of vehicle telemetry data to forecast component failures with 87% accuracy. Combined LSTM/GRU sequence models for degradation-curve forecasting with CNN-based feature extraction from multivariate sensor signals, surfacing real-time fleet readiness insights through interactive Streamlit dashboards.

GenAI & LLMs: LangChain, LlamaIndex, RAG Architectures, Multi-Agent Systems, Prompt Engineering, FAISS, ChromaDB, HuggingFace Transformers, OpenAI, Anthropic

Machine Learning: Time-Series Analysis (LSTM/GRU), CNN, NLP, Feature Engineering, Predictive Analytics

Backend Development: FastAPI, Spring Boot, Microservices, Docker, Kafka, REST APIs

Data & Databases: Python, SQL (MySQL, MSSQL), MongoDB, Vector Databases

• Finalist – Indian Army Innovation Contest 2025 (Top 20 nationally for RAG-based defense system)
• Top Performer – IIT Delhi RENDEZVOUS EN LIGNE (High-performance Python + Data Science solutions)
• GATE 2024 Qualified – 97.18 percentile in Computer Science
• Competitive Programming – 700+ problems solved | GeeksforGeeks Rank 120 (Score 1100) | LeetCode Rating 1599

M.Tech in Computer Science and Engineering
National Institute of Technology, Durgapur | 2024 – 2026
CGPA: 9.00

B.Tech in Computer Science and Engineering
Birbhum Institute of Engineering and Technology | 2018 – 2022
CGPA: 9.32

I'm open to collaborations on GenAI projects, LLM research, and challenging AI/ML problems. Feel free to reach out:

📧 chattopadhyaysanjoy18@gmail.com
💼 LinkedIn
🌐 Portfolio
📅 Schedule a Chat