Vaibhav Chhajed Email & Phone Number

Brooklyn, Ny, Us

San Francisco, California, Us

Us

Berkeley, California, Us

• Responsible for automation of data generation, model packaging, model deployment, model inference and monitoring.• Handled ML model deployment to production at scale, model changes, model monitoring, and generated alerts.• Developed environment for faster inference of ML model using TensorRT and Triton.• Build ML pipelines using Kubernetes, which replaced the manual work of creating an AI model and endpoint on GCP.• Build Data pipelines using Airflow, which automated the process and reduced the engineering work by 90%.• Build an Interactive Dashboard to track CV model accuracy daily to capture Data Drifts and Model Drifts.

Orange, California, Us

• Responsible for delivering end-to-end projects starting from PoC and taking them into an MVP phase.• Maintained reusable code for different ML pipelines to work efficiently.• Evaluated best MLOps work streams based on the project according to client requirements given time and efforts.• Generated recommendation engine to recommend podcasts and music on Amazon Personalize with 0.8 NDCG.• Discovered employee attrition trends based on data collection methods and filtered 40% of data.• Lead a team to develop an MLOps time series pipeline on AWS to forecast employee attrition with 80% confidence. • Developed a model to detect contaminants in garbage using a custom dataset with 65% accuracy.• Build end-to-end ML pipeline on Food classification and segmentation problems with accuracy up to 95%.• Improved accuracy by 2x through a regression model with more features over traditional rule-based models.• Guided juniors and freshers to get familiarized with ML tools and AWS services.

Pune, Maharashtra, In

• Managed multiple AI/ML projects for our clients in Media, Healthcare, BFSI and Legal domains.• Responsible for model monitoring and improvement • Created a better recommendation engine for 1M Indian regional movies and tv-series according to users survey.• Reconstructed Data Pipeline reducing 90% of redundant data for Hereditary and Reproductive Genetic Cancer.• Developed ML model to detect 15 different types of Hereditary and Reproductive Genetic Cancer.• Implement Chatbot which increased 2x productivity for booking appointments for clients using Google Dialog Flow.• Launched iZdox platform in a team for cognitive document processing being top 5 Startups in PuneConnect 2019.• Developed ML Sales intelligent platform which generates day-to-day signals based on the news on potential sales.

Master'S Degree, Computer Science

Bachelor'S Degree, Computer Science