As part of the watchdog platform team within the data science organization, worked on supporting addition of new anomaly types to the watchdog product, as well as redesigning the monolithic aggregation service that powered it.- Delivered the aggregate matcher service, as part of the redesign of the monolithic watchdog anomaly aggregation service. Responsibilities included prototyping, designing, implementing, planning and delegating work to 3 other engineers. The redesign provided a scalable solution for aggregating high volume anomalies of different types together.- Showcased company wide impact by ideating, prototyping, productionizing and supporting an internal slack app called Copydog with its own backend service. Copydog enabled users to easily extract and export information contained in slack threads elsewhere. The app currently serves ~6% of all Datadog employees every day.- Led another engineer to deliver a solution to enrich watchdog anomaly stories with data from the service catalog product. - Proposed and implemented an extensible framework for load testing services and catalyzed its adoption across other teams within the data science organization.- Mentored 2 engineers on the team, conducted ~100 technical interviews, presented at organization wide brown bags, conducted quarterly system risk reviews and fostered a culture of learning within the team by leading bi-weekly technical learning sessions.

As part of the media team within the catalog organization, worked on moderation of images, improving their quality and optimizing the performance of the media microservice.- Worked on various microservices on the Java, Scala, Python, Cassandra, Memcached, Elasticsearch stack.- Led the design, review, planning and implementation for a project that utilized the map-reduce paradigm to parallelize bulk media upload requests on the catalog platform at Grubhub. This increased stability of the media microservice and improved upload media times by 3X on the platform.- Delivered image moderation project for the media service to automate sanitization of all images being uploaded to the Grubhub platform, greatly reducing the need for manual intervention. - Implemented service level caching using Memcached hosted on AWS for the API that receives most of the traffic in the media micro-service, improving the response times by ~ 6 times. Thus also reducing the load on our Cassandra database cluster.- Collaborated with Site Reliability Engineers, Product Managers and UI/UX designers to design, implement and deploy the Image Quality Rating project for the media service using Amazon Mechanical Turk. This is essentially used to rate "how good" the food images shown to the diners are based on the lighting, plating and composition in those food photos. More than 300,000 images have been given a quality rating by the feature since its launch. - Enabled 5X improvement in the image upload times on the platform by introducing caching for the image upload API.- Reduced calls to the third party moderation service used by media service, by implementing a way to utilize existing moderation results for similar images.- Responsible for onboarding interns and new hires on the team. Mentored 2 engineers and conducted interviews to accelerate hiring.

As part of the search team within the AWS support organization, worked on measuring the quality of search results served to the support engineers.- Developed a utility to enable comparison of search result relevance for different search engines as a part of the AWS Support team.- Used this utility to prove that including only some domain-specific stopwords improves the relevance of the search system.

- Worked in the Data Science team for developing exciting,responsive and intelligent Data Science solutions for a Home Automation product.Major focus was on understanding the user's habits and preferences via multiple sensor data and using it to reduce electricity consumption of the household.- Also worked on the application backend development team to integrate Data Science solutions with the main application framework.

Developed a stand-alone utility in C on LINUX to assist an ongoing project for searching Call Detail Record files using multiple field searches and transferring them to another server.

Application development on Salesforce.com, a cloud based PaaS platform.