Dheeraj Madda
About Dheeraj Madda
Dheeraj Madda is a Deep Learning Engineer at RadiusAI in Bengaluru, India, specializing in optimizing neural network models. He has a strong educational background in Electronics and Telecommunication Engineering and extensive experience in machine learning and software engineering.
Work at RadiusAI
Dheeraj Madda currently serves as a Deep Learning Engineer at RadiusAI, a position he has held since 2023. Located in Bengaluru, Karnataka, India, he works on-site and focuses on developing advanced deep learning models. His role involves optimizing neural network models to enhance performance and efficiency.
Previous Employment Experience
Before joining RadiusAI, Dheeraj Madda worked at several organizations in various capacities. He was a Machine Learning Engineer at Avantari Technologies for six months in 2021 and later at Tiger Analytics for four months in 2023. He also held the position of Senior Analyst - Machine Learning Engineer at Tiger Analytics from 2022 to 2023. Additionally, he worked as a Software Engineer at MSys Technologies for ten months in 2020.
Education and Expertise
Dheeraj Madda has a strong educational background in engineering and technology. He earned a Master of Engineering in Electronics and Telecommunication Engineering from N B Navale Sinhgad College of Engineering, Solapur, from 2017 to 2019. He also holds a Bachelor of Engineering from the same institution, completed from 2013 to 2017. Furthermore, he obtained a Post Graduate Diploma in Internet of Things from the Centre for Development of Advanced Computing (C-DAC) between 2019 and 2020.
Technical Skills and Specializations
Dheeraj Madda specializes in optimizing neural network models to achieve reduced memory consumption, compact size, and low inference latency. He employs techniques such as Knowledge Distillation, Pruning, and Quantization. He is proficient in deploying machine learning models using runtime accelerators like ONNXRuntime, Intel OpenVINO, Nvidia TensorRT, and Nvidia Triton Inference Server. Additionally, he has experience in backend application development using FastAPI, focusing on scalability and performance.
Achievements in Machine Learning
Dheeraj Madda has made significant contributions to the field of machine learning. He achieved over an 80% reduction in training time for classifying new SKUs while maintaining accuracy for old SKUs through Domain-Incremental and Class-Incremental Continual Learning techniques. He also developed a custom trainer for distributed PyTorch training, which reduced development time by more than 10 times.