Product Graph

We use Machine Learning to search across Amazon’s selection, looking for products that might have errors and that may disrupt customers’ shopping experiences. Built models that utilize product embeddings to identify products incorrectly added to a family on the Amazon Detail Page.Building language models to identify entities in unstructured text.

Worked on many key features for my team. The stack included Java, Spring, AWS DynamoDB, Elasticsearch, AWS Lambdas, etc.

SocialCops is a data intelligence company on a mission to confront the world's most critical problems with data. It was founded because, even in today's data-driven world, no algorithm tells us the most optimized allocation of city budgets, no tool how to optimize medicine supply chains based on disease outbreak data. We partner with leading governments, philanthropies, corporations, and nonprofits to help them confront the world's most critical problems with data. I worked on Object Detection using the RCNN-based YOLO, YOLOv2 models, with Python being the language for code. Implemented pre and post-processing algorithms such as Histogram Oriented Gradients, Sliding Window, SVM.I also developed the Health Monitoring system using Grafana, InfluxDB, Deis Monitor and Prometheus. The published dashboard can be found here - https://grafana.com/dashboards/2577

I wrote the backend for this IOT-based startup right from the first line of code to deployment on AWS. Stack - Node.js + MongoDB