Projects
From distributed message brokers to self-correcting AI pipelines — I build systems that are production-ready.
Open to opportunities
Manas Sharma
Software Engineer studying Computer Science at Manipal, building distributed systems, AI/ML solutions, and production-grade applications.
Tech Stack
Beyond the Code
Jpop Music Reading Fiction Otaku Gaming Sports
Experience
•
Machine Learning Intern
Reliance Industries Limited
Built real-time occupancy detection using YOLOv11 and DeepSort across RTSP streams achieving 60+ FPS. Designed Multi-Camera Manager with spatial deduplication for 10+ concurrent feeds. Implemented adaptive tracking with motion blur detection, improving accuracy by 25%. Developed FastAPI backend with WebSocket real-time updates and SQLAlchemy persistence.
Projects Completed
From distributed systems to AI/ML pipelines,
here are some of the production-grade systems I've built.
AI Chair Occupancy Analytics
Real-time chair occupancy detection using YOLOv11 and DeepSort with multi-camera spatial deduplication, hitting 60+ FPS across 10+ concurrent RTSP streams.
Driving Scene Description Generator
An automated VLM pipeline that generates structured scene descriptions from autonomous driving images. Features 8 prompt engineering strategies, an 8-metric evaluation framework, and an AI agent for error analysis.
PriceScope
A production-grade multimodal deep learning system for marketplace price prediction. Trained on 1.48M Mercari listings with a BiLSTM + Attention + MLP fusion model achieving 0.430 RMSLE. Full stack with FastAPI, Next.js, and MongoDB.
Fleet Vehicle Data Gateway
A production-deployed real-time fleet monitoring platform with FastAPI, Next.js, WebSocket live updates, OAuth 2.0 authentication, and geofencing. Deployed on Railway + Vercel.
NeuralRAG
A self-correcting Retrieval-Augmented Generation system with LLaMA3, ChromaDB, and hybrid BM25+vector search. Features hallucination detection and automatic query refinement.
Distributed Message Broker
A high-performance, fault-tolerant distributed message broker built from scratch in Go with Raft consensus, gRPC transport, and topic-based pub/sub. Achieves 100K+ messages/sec.