Siddharth Mittal

Tax Testing Platform Team

Platform team that enables targeted regression and load testing for clients with millions of test case generated from production traffic.

• Designed and implemented the core replay functionality to replay production traffic against gamma stacks. Built a generic dispatcher to call a service endpoint and custom interceptors to extract replay context used to stub non-deterministic behavior in code.
• Worked on the high-level and low-level design for comparator service during re-architecture and migration to on-demand compute to solve noisy neighbor problem.
• Scaled the system to execute 100MM+ test cases every day across clients.

CoBranded Credit Cards (CBCC) Team

CBCC team is responsible for launch of co-branded cards in new marketplaces in EU region and provide a smooth customer experience as part of application process and credit card reward retrieval process.

• Designed the credit card application workflow orchestrators and implementation and testing of credit card upgrades in UK marketplace. More than 30,000 customers upgraded following this workflow
• Built a notification system to send emails to customers on changing preference for credit cards on Amazon
• Designed a platform to support installments on co-branded credit card transactions on Amazon. Implemented the critical component for integrating with banks to send transaction metadata like offers applied, bypassing issuer networks like Visa and Mastercard. Integration time reduced from 12 to 4 months
• Built a new 'Card Members' page on an Amazon proprietary platform. This had 10K+ daily customer interactions related to card benefits like reward points and updating customer preferences

Integrated Planning Platform Team

IPP team is responsible for creating planning platform, a single destination to meet the needs of all business teams doing Demand planning and Inventory Planning at Flipkart.

• Built a platform to power input driven planning across all of Flipkart by running statistical and ML forecasting algorithms written in R and Python at scale
• Worked on the capability to generate forecasts in near real time for individual verticals by updating inputs
• Designed and implemented functionality to view and compare Demand Plan vs Actuals data along any product hierarchy
• Platform used to forecast monthly and quarterly demand for 10MM+ products on Flipkart

Custom Branded Experiences Team

CBE team is responsible for creating sponsorable products for use in Yahoo. Worked on developing reusable web modules for advertisers ensuring performance, scalability and consistency across multiple browsers and devices.

• Photoheader Module: The most extensively used sponsored module in Yahoo ecosystem generating more than half a million views and about 100K actions on a daily basis.
• Countdown Module: A simple countdown module for events like Oscars and Grammys was developed from scratch. Implemented code in Javascript to resolve issues like client-side lag and pjax calls.
• Carousel Module: Worked on a sponsorable module with carousel functionality and Tumblr as data source. The module uses Tumblr apis to extract data from blogs and jQuery for the DOM manipulation

Multicasting Team

Multicasting team was responsible for enabling multicast functionality using Juniper's proprietary markup language

• Created a test suite of smoke tests for 'Automation of Multicast Snooping'. Covered test scenarios for all possible bridge and interface combinations.
• Got return offer for full time job at the end of internship.