Hi, I'm Shubham Kulkarni 👋

AI Engineer · Computer Vision · Generative AI · RAG · Edge AI · MLOps

1+ year of professional experience building production-grade AI/ML systems.

     

I'm an AI Engineer with 1+ year of industry experience, currently working on production Computer Vision, Edge AI, and Intelligent Transportation Systems.

I build AI systems across the complete lifecycle — from model development and optimization to real-time inference, RAG, vector search, AI agents, APIs, and production deployment.

• 🚀 50K+ real-time events/day
• 🎯 95% Computer Vision accuracy
• ⚡ Sub-40ms processing on NVIDIA Jetson
• 🔍 30× faster vector retrieval using FAISS
• 🚗 150K+ forensic queries/day
• 🏆 AIR 1 — Smart India Hackathon 2023

AI / Computer Vision

Python · PyTorch · TensorFlow · OpenCV · YOLOv5/v8 · RT-DETR · ByteTrack · OSNet-AIN · PaddleOCR · Hugging Face

Generative AI / RAG

LLMs · RAG · LangChain · FAISS · BM25 · Hybrid Search · RRF · Agentic AI · Prompt Engineering · Knowledge Graphs

Edge AI

NVIDIA Jetson · TensorRT · CUDA · ONNX Runtime · GStreamer · RTSP

Backend / MLOps

FastAPI · Django · Flask · Docker · Kubernetes · CI/CD · PostgreSQL · Redis · REST APIs · WebSockets

Cloud

AWS · Azure · Google Cloud · Vertex AI

Multi-agent AI platform for smart-city decision making, combining AI agents, digital-twin simulation and explainable AI for risk-based recommendations.

Tech: Gemini · Vertex AI · FastAPI · Next.js · Docker

🔗 View Repository →

Local-first document intelligence platform for PDFs, images and web content, featuring hybrid retrieval with FAISS + BM25, RRF, knowledge graphs and LLM-based Q&A.

Tech: Python · FastAPI · FAISS · BM25 · Mistral · LangChain

🔗 View Repository →

Real-time vehicle re-identification system using YOLO, OSNet-AIN embeddings and FAISS IVF for high-speed search across surveillance data.

150K+ queries/day · Sub-100ms retrieval · 30× retrieval improvement

Tech: PyTorch · YOLO · OSNet-AIN · FAISS · FastAPI · Docker

🔗 View Repository →

Agentic AI system that correlates IoT, camera and audio signals to identify anomalies, validate incidents and generate automated safety responses.

Tech: FastAPI · Agentic AI · RAG · Vector DB · Docker

🔗 View Repository →

AI + IoT platform combining smart irrigation, crop-disease detection, predictive analytics and multilingual AI assistance.

🏆 Rank 2 — SPPU Startup Olympiad 2025

🔗 View Project →

Arya Omnitalk Wireless Solutions Pvt. Ltd. · Pune

Aug 2025 – Present

Building production AI systems for Intelligent Transportation Systems.

• Architecting Computer Vision pipelines processing 50K+ events/day
• Deploying and optimizing AI inference on NVIDIA Jetson
• Using TensorRT, CUDA and GStreamer for high-performance edge inference
• Working with YOLOv8, RT-DETR, ByteTrack and PaddleOCR
• Building asynchronous backend and real-time alerting pipelines
• Developing vehicle re-identification and forensic search systems
• Supporting production systems with 95% CV accuracy and 99.9% edge-node uptime

🥇 AIR 1 — Smart India Hackathon 2023 National Champion · Software Edition

🥈 Rank 2 — SPPU Startup Olympiad 2025 JalTantra · AI-powered AgTech platform

🥉 Global Runner-Up — TIAA Global AI Hackathon AI Investment Intelligence Platform

Bachelor of Engineering — Electronics & Telecommunication

Savitribai Phule Pune University Dr. VVP College of Engineering · 2021 – 2025

First Class

• Deploy Multi-Agent Architectures — Google · 2026
• Introduction to Machine Learning Concepts — Microsoft · 2026
• Engineer AI Agents (ADK) — Google Cloud · 2025
• Career Essentials in Generative AI — Microsoft + LinkedIn Learning · 2024
• Deep Learning Onramp — MathWorks · 2024
• Statistics Onramp — MathWorks · 2024
• Python Programming — GUVI · 2023
• Geodata Processing using Python — ISRO · 2023

I'm interested in AI Engineering, Computer Vision, Generative AI, Agentic AI, RAG, Edge AI and challenging real-world AI systems.

   

Building AI systems from models to production. 🚀