Amazon SageMaker

Amazon GO

Interned with the ImageX team at Uber that specializes in using Computer Vision to enable localization for enhanced user experiences.

Evaluated the impact of filtering matches between 3D point clouds and query images based on semantics.

In order to get Dense 3D models from sparse 3D point clouds (from SfM), I evaluated Depth Estimation and Depth Completion techniques using Deep Learning to get high fidelity depth maps.

For the task of Depth Completion, I generated sparse depth samples from sparse 3D point clouds by projecting 3D points into a 2D image space.

As Senior Software Developer, I was responsible for suggesting and implementing changes at an architectural level.

This led to the use of Redis for caching instead of making frequent requests to internal microservices, use of a publisher-subscriber model to replace a polling architecture, and building components for dynamic form generation.

As part of the Manufacturing Software Team, I was responsible for the design and development of internal web applications for Vistaprint's Manufacturing Units. Working with JavaScript Frameworks such as Angular, Vue.js in an Agile environment, I have developed a strong foundation in Web Development. I was responsible for building microservices in GoLang and Node.js. Having deployed applications to AWS familiarized me with services such as EBS, S3, Load Balancers, EC2, ElasticCache, DyanmoDB, SQS/SNS, CloudFront.