• Building a novel graph database-based semantic search engine using Neo4J and Python to create a platform to search for experiments based on the published data generated on Texas Advanced Computing Center (TACC) supercomputers. This improves the reach of experiments and reduces duplicated effort in terms of time and resources.

• Developing a CI/CD system that leverages Python and the Kubernetes computing clusters to perform continuous integration on applications deployed on TACC. (tuitus-ci.github.io)

• Developing a differentiable Material Point Method (MPM) simulation library in Python using JAX to use MPM models in neural networks.

• GSoC '23 mentor for diffmpm.

● Created a Looker dashboard to monitor the health of various business metrics and data preprocessing pipeline setup and improved the working timelines by reducing monthly effort by 20%.

● Drove cross-functional team discussion and analysis of different features' importance on ML model performance, and changed the decision by showing a 2% metric improvement.

● Created a Looker dashboard to collate data for different key metrics for the project, and monitor the overall daily performance of the product.

Working on large scale epidemic simulations and using Machine Learning techniques for Inverse Problems