◦ Led the development of a simulator that leverages ERP data to create a digital twin of the supply chain, allowing for the simulation of operations and the evaluation of AI-based decision-making against traditional control systems
◦ Applied Reinforcement Learning to automate inventory management across diverse industries, including distribution, manufacturing, and oil & gas supply chains, while boosting operational efficiency and cost savings
◦ Optimized ML workload throughput by refactoring and enabling horizontal scaling using Ray clusters on Google Cloud and Microsoft Azure, resulting in faster training times and improved model performance
◦ Leveraged Docker and Kubernetes for hyperparameter tuning, Tensorboard logging, CI/CD workloads, etc.
◦ Built and orchestrated data pipelines using Apache Airflow and (Py)Spark for ETL and ensured high data quality through a custom data unit testing framework designed using Deequ library in Scala
◦ Maintained backend services using Ruby on Rails, Node.js, and MySQL to support data-intensive REST APIs
◦ Developed a client-facing dashboard with Vue.js, providing custom charts and metrics relating to supply chain KPIs
