Shriram Jugnake
Assistant Manager at Tata Consulting Engineers Ltd. - Deep Learning, Image Processing, Machine Learning
Mumbai
5+ years of work experience
Jun 20251 yr 1 moPresent
Navi Mumbai
Digital Twin & Physical AI: Architected multi-domain digital twin platforms for industrial environments including aluminum scrap yards and high-speed railway stations using NVIDIA Omniverse and Isaac Sim. Built Python extensions within Omniverse Kit to control USD prims, real-time 3D viewport widgets for interactive asset overlays, and a custom connector to integrate IBM Maximo enterprise asset management data with interactive 3D USD models. Handled full 3D pipeline — importing FBX/Revit models, converting to USD, and publishing to Nucleus for team collaboration. Deployed Web Viewer (WebRTC) for real-time streaming and multi-user collaboration.
Agentic AI & RAG: Built an AutoGen-based multi-agent AI system for asset impact analysis, root-cause detection from datasheets, and recommendation generation from manuals using real-time MQTT sensor data. Built a hybrid RAG pipeline using NVIDIA NIM Nemotron parser, Llama, and vector databases with a VLM for engineering drawing parsing. Optimized LLM costs by deploying smaller local LLMs with refined prompt engineering for faster inference. Integrated HMI and KPI monitoring into the digital twin environment.
Computer Vision & Video Analytics:Developed segmentation PoCs for X-ray imaging, concrete pore detection, and steel surface dent inspection using Mask R-CNN and U-Net. Built an AI platform to process 6K+ engineering drawings using Layout Parser and Qwen-VL for structured data extraction.
Robotics & Simulation (Exploration): Explored ROS2 integration with Isaac Sim — controlled an articulated forklift via ROS2 topics with physics joints and drives for multi-axis motion simulation.
Leadership: Led a 4-person team delivering a real-time thermal plant asset monitoring portal. Owned architecture, planning, and end-to-end delivery from R&D to client demos.
Nov 20231 yr 8 mosJun 2025
Mumbai, Maharashtra, India
Oct 20212 yrs 2 mosNov 2023
Mumbai
Developed and deployed production-grade computer vision systems for industrial quality inspection, with end-to-end ownership from dataset collection and model training to hardware integration and on-site client delivery.
Diamond Grading System
Built an automated image processing pipeline using FLIR camera SDK to capture multi-angle facet images of diamond stones. Developed Python and OpenCV algorithms to quantify optical properties — fire, brightness, and scintillation — for automated grading output. Engineered precise multi-axis hardware positioning to systematically capture every facet angle required for accurate inference.
Diamond Authenticity Classification
Collected and annotated real-world datasets on-site at client facilities for multi-class stone classification under UV and fluorescence lighting conditions. Trained Faster R-CNN and Mask R-CNN models for stone detection, instance segmentation, and authentic vs. fake classification. Integrated the full inference pipeline with live FLIR camera capture — real-time frame capture → model inference → per-stone classification → automated operator output. Delivered complete end-to-end deployment on production hardware with on-site installation and client validation.
Key Technologies: Python · OpenCV · Faster R-CNN · Mask R-CNN · FLIR Camera SDK · Instance Segmentation · Multi-class Classification
May 20212 mosJun 2021
India
Internship on #Automation
Education
Bachelor of Engineering - BE, Mechatronics Engineering
Terna Engineering College
20174 yrs2021
Join Shriram and 10k+ other professionals on
Tal is the app where bosses meet engineers & builders directly, without middlemen
Download Tal Boss