Designed, architected and led the development for the airtel mobile site and app, along withthe first implementation of the airtel mobile money.

Reporting and analytics reporting service leveraging Diageo Insights with datawarehousing for Asia Pacific

Custom built an end to end Data platform starting all the way from Data ingestion and consolidation, reporting, predictive analytics and dashboarding & Visualization. Built the team of 15 members consisting of data scientist, big data engineers, busness analysts and ML engineers, personally arhitected the solution, custom coded the data engine. built a query bot for custom queries on over 200 paramaters. Addressed domain issues like consumption patterns, distrubtor sales channel performance. SKU performance, distribution and marketing optimization. Recommendation engine.

Income tax department of India, wanted to improve the efficiency and accuracy of grievance redressal. They also wanted the level of MIS and analytics that it provides its stake holders. We introduced significant improvement of over 85% within a short period of 4 mounths. We built a custom nlp engine which read the complaints and grievances of the end user. It identified the core issues and helped serve the right resolution to over 75% of the cases at the first point of contact. another 17% were saggregated and relevant data was pulled including other remote data sources, historical resolutions and patterns and served to the subject matter expert to endorse and hence helped them resolve the case at the earliest. Only 8 percent of the cases required a more detailed intervention due to dependencies on unavailable data sets which were outside the purview of the engine.

The resulted AI based solution during the pilot implementation helped get the grievance redressal rate to over 95% within the same week of reporting

Developed a NLP engine that could be used to read the free text input from the customers, describing their symptoms and problems. The engine analyzed the reported symptoms and used its proprietary technology to recommend the most apt clinical tests in real time. This helped to reduce over 30?lls to the physical call centre. Within 72 hours of deployment the tool helped the client close 36% more orders. The second stage was the development of Osler chatbot which worked to replace the complete backend call centre. The product had free speech readibiity with extremily high readability. It used NLP and NLG to respond in real time. It was built on a completely custom built NLP engine.

The client approached us to build a inventory management solution for their medical supplies unit, the site was to work with a very heavy dynamic user management system, an extremely complex inventory, supplier and predictive ordering system. It had a custom reports section with insights and recommendations based on the usage patterns. It also had an internal recommendation system

Worked as a Data and systems analyst on one of the largest nationwide Clinical Health Records Project at the time, built the pipeline and remote analysis mechanism back in 2003. Deviced the mechanism for quick access to PACS and other patient records at the click of a button over a private vpn network.

Built the statewide health management information system for an ADB (asian development bank) funded project, from scratch. The uniqueness about the project was that it was made for situations where we could transmit data over regular PSTN lines which were intermittant in nature, but still ensure data integrity. The project was nominated for the National E governance award. It was the first attempt that connected over 14000 Asha (Ground level Hath workers) PHC(Primary Health Centres), CHC ( at Block Levels) and the Central station, it reduced the transmission and reporting of data from 45 days to less than 2 days. There by ensuring quick action in the event of an Epidemic.