Lead the Data Science and Machine Learning vertical for Zomato Online Ordering vertical. I oversee the search, personalization, user journey, customer segmentations, discount optimization, fraud detection and ad revenue optimization among others. Created multiple solutions which are now integral part of our business and have resulted in increased profitability across many areas.Been instrumental in embedding our solutions across various customer touch points and inculcated the 'product first' thinking in all our solutions. You name a problem at Zomato and I am pretty sure we have had a fair bit of impact on solving it.

Leading the Search and Personalisation data science vertical. Responsibilities include - improving the order conversion per app open and also improve the unit economics of orders through personalisation.In past created purchasing power segmentation for online ordering platform which helped us create a segmented level targeting and subsequent reduction in promotional burn per order.Created dish level entities for the online ordering platform and improved dish search conversions

• Led an omni-channel (web + mobile app) product recommendation engine for a QSR chain. - Real time product recommendations during basket checkouts. - Incremental uplift of 5% in additional sales compared against existing recommender system.• Engineered a variant pricing algorithm and helped an automotive giant re-launch their mid-range SUV through a revisited price positioning within the market. 3 months since launch, sales are up ~20% QOQ.• ‘Demand Pools’ and ‘Need States’ based customer segmentation for pharmaceutical company to acquire, retain and maximize customer engagement.• Creation of a machine learning hands-on course content for the team.

- Own roadmap to develop scalable data science products to aid client’s long term business strategies.- Designed effective digital media display campaigns using customer targeting models through personalized campaigns and impactful displays. - Automated targeting models for 15mn customers; brought down project lifecycles by ~50% and saving 2 FTE’s- Created customer ‘look alike’ models; helped business seal a deal of $200K with leading social media partner- Bring all our measurement capability under one suite which leverages cutting edge science to evaluate the performance of all customer 'engagement' channels for all our retail and FMCG partners.Measure the effectiveness of digital 'personalization' suite used on retail partners websites. - Evolve our measurement sciences using latest methodologies to expand our solution offering.

- Created customer targeting model on 2 million shopper sample to help major FMCG client win new customers for its newly launched product and realize +£4mn in sales (>40% growth). //Scaled up the solution for other FMCG partners, potential of ~£1 million for dunnhumby in revenue.- Achieved 99.4% accuracy on MNIST data and trained India analysts on Image Classification using CNN’s.- Conceptualized product recommender tools using collaborative filtering and predictive modeling for brands.- Overhauled the global segmentation of customer ‘shopping missions’. //Conceptualized the idea of establishing transaction level ‘product interactions’ on millions of shopping transactions; clustering those ‘interactions’ to identify various ‘missions’ across 10 countries. //Empowered retailers place relevant product range in each store inspired from these missions.- Led product development of a query based reporting engine feeding into ~20 global reports and helped analysts cut down code development time by 50%. Currently a dunnhumby global solution.

- Co-created the science based ‘inflation’ module and globally rolled out the customer loyalty segmentations.- Designed and implemented the robust engine to account for macro economic conditions and thus arrive at the optimal criteria to classify customers based on their spend.- Leveraged 'product substitutability' as the concept on web browsing data to understand customer's browsing behavior and determine product associations to help store managers with in store product displays- Challenged the current set up defining 'price important' products. Proposed and tested few hypothesis and offered insights which eventually facilitated the formulation of strategy and established best practices around defining 'price important' products.

- Creating the Incentive compensation plan for medical representatives of a top pharmaceutical company- Responsible for creation of call plans for medical representatives

Working with the Mortgage Loan Origination System (LOS) Team in Tavant Technologies. Associated with the development and integration of various modules of the LOS.