• Tech Lead for MoPub Ad serving. Lead for Adserver and several auxillary services that supported MoPub Adserving. C++ backend service received > 2M requests/second and powers MoPub SDK embedded in several popular apps.

• Proposed, Architected, Designed and Developed MoPub Experimentation platform. Experimentation platform allowed us to run several important experiments on Adserving and Auction Marketplace when launching new features and optimizing existing ones. This involved putting together a proposal and getting buy-in, leading design and execution across backend, front-end and data teams to ship the platform. Shepherded experiments in making sure experiments being proposed and run were feasible and correct.

• Lead design and development for Advanced Bidding, which opened monetization from Ad networks in addition to DSP's and allowed MoPub to run unified auctions. Post GA, Advanced Bidding was integrated into all of the managed publishers at MoPub.

• Lead development for Split-net based creative early filtering for MoPub in Twitter's Ad Auction. This improved Twitter's revenue from MoPub by +1.31%, while improving click installs by +2.75% from MoPub and improving advertiser ROI by -1.37% cost per install.

• Lead development for introducing new Ad Format in MoPub - Rewarded Ads.

• Reduced late drops for Twitter bids into MoPub by 85%.

• Reduced infrastructure costs by 4M/year by switching data stores from Manhattan (Key-Value store) to Nighthawk (Redis-backed store) for user frequency capping and adding a rate limiter for Ad requests.

• Migrated pSocial model learning to GCP from on-premises. Also improved model's CTR by 6.9% and Engagements by 8.9% via updating feature selection, tuning traffic and adding display correction.

• Mentor several engineers across teams. Conducted onboarding sessions and classes for introducing MoPub's backend architecture.

• Evaluate proposals and guide product and management on feasibility and estimates.

• Architecting and Leading backend design and development for large-scale features and services that power the With Friends network of games at Zynga. Ruby on Rails backend service powers Words With Friends, Boggle, Crosswords With Friends and other With Friends titles with >2M CCU and >10M DAU.

• Launched Words With Friends 2, which reached #1 in Top Free Apps on iOS and Android App

stores.

• Designed and Developed real-time game Stats for With Friends games that process 20M

moves/minute. It replaced an existing Stats system that was expensive to run (>1M/year) and

only updated once a day, whereas the new system costs ~100K/year and is real-time.

• Architected, Designed and Developed a Cross promotion system for the With Friends games, where a game in the network can surface relevant game information/reminders in another game and rewards players for taking actions in a cross-promoted game. This helped to launch 2 new games Boggle and Crosswords With Friends by providing a steady funnel of installs and DAU from Words With Friends and other With Friends titles.

• Designed and Developed the Challenge system that powers Events, Weekly Challenges and Daily Goals for the With Friends games. The Challenge system records and serves progress for all moves and multiple in-game actions in real-time.

• Designed and Developed Discover, a players graph network for the With Friends games that surfaces a stream of players that are likely to be a good match for the user.

• Redesigned and Fixed Community Matchmaking service, a central social feature for New Words With Friends using Scala, Akka, Spray, ElasticSearch, and MySQL.

• Designed and Developed back-end micro-services using Java, Scala, Akka, Spring, MySQL and Redis for an unpublished product.

• Perform architecture design reviews, code reviews, and breakdown product specs.

• Lead and mentor engineers in the team in shipping high-quality code without slipping deadlines.

• Developing Indigo’s flagship product Ascent, a web-enabled automated chromatographic analysis, and review system.

• Implementing features for Ascent using Ruby, Rails, MongoDB, RSpec and Cucumber with TDD/BDD.

• Developed WCF services to support workflow between instrument machines and Ascent servers.

• Designed and Developed a method editor that helps scientists configure assays in a sandbox and deploy them into production.

• Optimized mongo queries and data transfer points to achieve 20,000% improvement in data review speed in Ascent.

• Designed and Developed a configurable data dealer application that is part of the Ascent workflow and is installed on the instrument machine as a windows service to process and filter raw data.

• Implemented the Dealer application using C#, NUnit, and SpecFlow.

• Developed and extracted an API in C# that is being used by the data dealing and data reduction services in Ascent.

• Implemented features for a quality rules API framework that allows customers to create rules for assays in Java, JUnit, and Mockito.

• Developed features for a service that consumes data from other services to validate them and transform them into mongo documents.

• Developed a standardized sequence file generator for Ascent using C#, MSpec, and SpecFlow.

• Added features to the data reduction service that is part of Ascent using C#.

• Developed expertise in test driven and behavior driven development.

• Developed Jenkins jobs workflows for various components of Ascent.

• Pair programming and mentoring other engineers in writing clean, testable and maintainable code.