Aditya Chaudhry Email & Phone Number

San Francisco Bay Area

• Trained and deployed object detection models for defects in Transmission and Distribution powerlines, reducing data processing cost and time by 50% for customers
• Developed a Machine Learning workflow management system with ClearML, automating model training processes and reducing manual effort by 50%, while lowering GCP costs by 25% through the implementation of an autoscaler
• Automated model debugging and utilized image embeddings, enhancing active learning efficiency by 40-50%
• Curated and optimized datasets for model training, ensuring high-quality data selection to improve model performance and accuracy
• Conducted Exploratory Data Analysis (EDA) on data collected from IoT-enabled drone imagery to gain deeper insights into data patterns, which informed feature engineering and improved model development

Champaign, Illinois, United States

• Facilitated the migration of the Hadoop Analytics project (Starling) to AWS, implementing a streamlined transition plan that resulted in a 20% improvement in data processing efficiency. This included re-architecting.
• Developed and maintained Python scripts for AWS Lambda to automate log parsing tasks, which reduced manual effort by 40%. This involved designing robust, scalable scripts to handle various log formats and integrating.
• Devised a trigger mechanism for efficient data transfer to AWS databases, enhancing data flow speed by 10%. This solution involved creating custom triggers and workflows to ensure timely and accurate data.

Mumbai, Maharashtra, India

• Engineered a comprehensive leave management web application that streamlined the leave approval process, improved workflow continuity, and reduced the time spent on leave-related administrative tasks by 30%
• Coordinated cross-team collaboration to manage ETL (Extract, Transform, Load) testing and data integration efforts, which led to a 25% increase in data accuracy. This included facilitating communication between teams.

New Delhi, Delhi, India

• Developed an optimized Image Classification model for Military Aircrafts using Keras, CNNs, and Transfer Learning
• Implemented Data Augmentation and Regularization techniques to boost model accuracy from 77% to 91%
• Curated and preprocessed a high-quality image classification dataset, improving data consistency and optimizing model training efficiency, leading to enhanced overall model performance
• Collaborated with team members at Defence Research and Development Organisation (DRDO) to achieve project milestones

Master Of Science - Ms

Completed my Master’s degree with a specialization in Machine Learning and Big Data, involving hands-on projects that included developing.

Bachelor Of Technology - Btech, Information Technology