Renga Pragadeshwar
@rengapragadeshwar
Data Scientist - Manager at Ohmium Operations
Bengaluru, Karnataka
Renga Pragadeshwar is a Data Scientist and Manager at Ohmium Operations with a background in Mechanical Engineering from IIT Madras. He has a proven track record in developing machine learning models, including LLMs and computer vision systems, for the automotive and energy sectors. His work at Jaguar Land Rover led to significant cost savings through anomaly detection and process automation. Renga is skilled in Python, TensorFlow, and cloud technologies.
Experience
Data Scientist - Manager
Ohmium Operations
Curated and pre-processed a diverse corpus of market-related textual data, including financial news articles, social media sentiment, and analyst reports, to train the GPT-3 based LLM. Designed cost optimization and guarantee delivery strategies using Machine Learning and objective programming. Crafted a pioneering Levelized Cost of Hydrogen (LCOH) calculator with Flask, transforming it into a web application.
Data Scientist
Jaguar Landrover
Recommendation system for quick access of remote feature - deployed using AWS Sage maker into production. Developed a vehicle performance metrics data visualization dashboard using Tableau. Developed a comprehensive bench-marking framework for evaluating speech systems in voice recognition applications. Implemented a regression-based model to preheat batteries, effectively reducing charging time by 30%. Implemented anomaly detection in battery signals using range filters and distance metrics, leading to savings of 350 million GBP. Led the upgrade of invalidated HMI to automated validation utilizing YOLO v4 for icon detection and CRAFT/tesseract powered text recognition. Designed and implemented a DeepLabv3+ encoder-decoder model for semantic segmentation of road features in ADAS. Developed, integrated, and deployed health checker API for fedora - Linux based Infotainment stack.
Education
Indian Institute of Technology - Madras
Dual Degree (B.Tech & M.Tech)
Mechanical Engineering
Thesis: Developed a cost-effective 3D tactile sensor for ocean applications, enabling temporal and spatial measurements with 1mm sensitivity and high-pressure range. Utilized OpenCV for data processing and visualization.