Apr 2019 – Present · 6 yrs 11 mos · Greater Bengaluru Area

• Dailyhunt has built and operates one of the most sophisticated publisher owned ad recommendation and serving system in the country that is constantly getting upgraded. This team serves ads with real time performance (click, install, leads etc.) prediction at ~70,000 RPS hitting the servers, checks with dozens of ad networks along the way, and still responds with the most optimal ad for over a billion requests daily. As a leader of this team, I have had the opportunity to reimagine data pipelines, create new recommendation systems for products, generalized automated classifier creators for clicks, app installs, leads and eventual transactions for direct and network sourced campaigns. As a revenue maximization system this business critical system has been built to optimize for complex objectives for the benefit of advertiser and revenue and user experience. All of that is achieved by a large number of simultaneously operating ML models while self correcting for delivery changes every few minutes by absorbing the >50K RPS stream of events from the Android, IOS and PWA applications. To finely play this balance we have built realtime analytics at scale for business insights and sales planning across objectives of reach, inventory & revenue. To extend the goodness of the platform to other publishers and networks, we have also recently built a Vernacular Audience Network that operates over RTB and connects the regional language publishers with advertiser demand.
• Always on the lookout for great product engineers who consider their yesterday's winning work as not good enough per their current understanding. Temporary pause of hiring across levels for backend, ML, data & fullstack.

May 2018 – Feb 2019 · 9 mos · Greater Delhi Area

• The primary responsibility of this role was to lead the technology team of the DHISCO hospitality switch that joined Rategain as M&A deal in 2018. Integration of cultures, teams, processes and systems between the Dallas and Noida teams while meeting synergy goals for the acquisition leading to over a million dollars of OpEx saving. Revival of the roadmap to be able to address the hybrid market between the channel managers and switches in the hospitality industry's distribution of Rates, Inventory and Availability in near realtime.

Sep 2014 – Mar 2018 · 3 yrs 6 mos · Noida, Uttar Pradesh, India

• As a director of engineering, Ajay leads the Machine Learning/AI, User Experience, Location Intelligence Data Engineering and Geocoding Engineering teams at Pitney Bowes. He has played key role in the PB transformation journey energizing teams by championing initiatives involving automation and machine learning around data, and geocoding rapid global expansion. Driving market-focused innovation and efficiency leading to recent product launches like Master Location Data, Geo Enriched Data and many others, he is also taking the teams’ skills and consequently sustainable engagement to new highs such that they are able to do the new products while delivering existing commitments.

Jan 2013 – Sep 2014 · 1 yr 8 mos · Greater Hyderabad Area

• 1. Web ranking and relevance work for the local vertical of Bing in a google compete scenario. Our results are close to the leading search engine in side by side comparison, and we continue to improve them at a faster pace than the percieved leader by use of various traditional ranking and machine learning techniques.
• 2. Managed the development of one of the two new Bing branded applications bundled with Windows 8.1. We built this application on Windows 8 platform along with services built on top of Bing, MSN and Azure stacks of Microsoft that scale to 100s of millions of users. Our applications are also designed to be role models for the Windows 8 app ecosystem. Responsible for architecture, build, deployment and availability of the app and it's associated online services.

Sep 2011 – Jan 2013 · 1 yr 4 mos · Bangalore - India

• 1. Managed one of the two teams responsible for Yahoo! Home Page Engineering.
• 2. Integrated simple homepage markets in the new media stack of Yahoo Publishing Platform
• 3. Launched Yahoo! homepage for a new european market leading to over 90% market share.
• 4. Managed Yahoo Games! team as part of the media organisation.

May 2009 – Sep 2011 · 2 yrs 4 mos

• Staffed up an engineering team to design, develop, operate and sustain over dozen digital media properties on the web and their mobile versions. Set up performance metrics and processes to align the team.

Oct 2006 – Apr 2009 · 2 yrs 6 mos

• Class Grid was an education sector startup with an aim to make education relevant based on big data analytics from previous years. It would create a better instructor - student pairing while providing a platform to both share educational content and measure it's effectiveness in learning.

Mar 2006 – Sep 2006 · 6 mos

• Staffed up an engineering team to design and develop the workflow and eCMS integration connectors for Adobe LiveCycle with the popular ECMSs, Documentum, Filenet and IBM CM.

Aug 2004 – Jun 2005 · 10 mos

• Consulting as part of the Global Architecture and Core Technologies Group. Specialized in Web Services and SOA. The two assignments were in large scale systems in Investment Banking and Government verticals.

Sep 2000 – Aug 2004 · 3 yrs 11 mos

• As a member of the J2EE Application Server group, I was responsible for a large number of subsystems and components. Also led to two dozen patents.

Jun 1999 – Aug 2000 · 1 yr 2 mos

• Research and development of various subsystems of the Mirim product including a Bayesian Belief Networks module for cross-sell and up-sell predictions.

Jan 1996 – Dec 1997 · 1 yr 11 mos · Chandigarh, Chandigarh, India

• As a member of the student body, was involved in raising multiple international internship opportunities with the local businesses. These internships allow overseas talent to take up upto 6 months of traineeships and help companies expand to other geographies after they leave.