At SoFi, I played a pivotal role in API development, leveraging technologies such as Python, AWS, and Docker. My contributions included leading the BillPay Switch project and creating essential APIs like the Token Service and Events API. I also enhanced backend services for Wire Transfer and ACH Transaction Simulation, focusing on security improvements.

Used Azure cloud to build solutions for customers.

Completed Azure certification (DP 200) and PowerBI certification

Interacted with clients to understand their needs and provide useful solutions and insights

Fellow, Insight Data Science, San Francisco

 Created DevOps pipeline for databases by using EKS,ECR, CircleCI ,SQLAlchemy and PostgreSQL to manage database from

development to QA and from QA to production

 Implemented the migration of databases with traceability to database changes and ability to roll back by creating tests in dev, QA and

production environments

Responsibilities : Responsible for the regular data analytics and report generation of key performance metrics, assisting project teams with one off tasks when available, Assist with Continuous Delivery

Amazon Web Services, East Palo Alto : Software Development Engineer Intern (TensorFlow team) May-August 2019

1. Worked on SageMaker-Debugger, GitHub link- https://github.com/awslabs/sagemaker-debugger

which helps to analyze the tensors generated during training job of neural networks.

2. Created Index Writing class for that project so that the tensors that are being saved for analyzing can be directly fetched from the exact location it is saved. Also, created utility functions that can later be used to fetch files for exact step, integrated Index Writer with Index Reader. The speed to fetch tensor improved dramatically because of Index Writing.

3.Fixed some bugs for the project, like:

Detection of end of training job, raising run time error if same directory structure is being used, Updating tests for more generalized usage.

4. Created Code Integration/Code Deployment system to run all tests for each PR and upload reports and wheels for tool to S3, which can –

Run integration and unit tests for each Pull request, Publish the results of Pull Request to the corresponding Pull Request, create pip wheel package from alpha and master branches, publish pip-packages to corresponding s3 locations if tests are success. Build sends chime notification using lambda function to deep engine group about warnings /errors and info generated during the build as well as link to build logs.

4. Learned a lot about TensorFlow and about AWS tools like CloudWatch, lambda, Codebuild, Codepipeline, S3.