Part of the Data Infrastructure team which manages Facebook’s PLANET scale, multiple ExaByte size Data Warehouse . Lead Data Quality: Bad data has far reaching consequences. The data computation is a graph and data quality issues for one table impact all the multiple levels of downstream tables which depend on it. - Proposed and led the team to implement an anomaly detection based automatic data quality framework which took quality coverage from 11% to more than 80% of the important warehouse tables. Lead, Data Ingestion: The team builds and operates the infrastructure to bring 10s of PetaBytes of data daily into the data warehouse. These systems have to be optimized to fetch data at scale and have extremely high reliability. - Worked with the team to define the next generation architecture for the dimensional data ingestion.

Engaged with issues of privilege, oppression and bias in our schools and society. Also dived deep into research around learning science and motivation.Closely studied a lot of technology solutions for education and used many of them in my own classroom. Also talked to a lot of stakeholders - teachers, students, administrators, researchers, VCs, Ed Tech companies - to build deep insights about technology use in education. Most products have a core idea that adds value. But they often underestimate the effort required to integrate their solution into a classroom environment. In order to be successful, using a product should give time back to the teachers and not take it away!

ThoughtSpot is building a revolutionary product to disrupt the Business Intelligence market. Check us out at http://www.thoughtspot.com.PRINCIPAL ENGINEER (May'15 - Nov'16) Contributed to all areas of the technology stack. Besides leading design direction and technical decisions, this included writing the first data model for the backend in-memory database, writing the search indexing algorithm which helped take us to market and leading the monitoring, alerting and analytics charter for the company.HEAD OF ENGINEERING (Jun'12 - May'15)As a cofounder and head of engineering, helped build the engineering team and the product from ground up. Also worked with other business functions such as sales, marketing and operations to bring the product to market and successfully deploy it with both Fortune 500 and smaller customers. Helped the company grow to 50+ people.

Lead fact extraction and data normalization for content pipeline at Amazon. Worked in the area of extraction of structured information from unstructured or semi-structured data. As part of this, managed three teams focused on - Applying data normalization to 100s of millions of products on Amazon - Using clustering and data mining algorithms to extract canonical representation (such as brands) from ALL amazon products and mapping the incorrect values to the canonical counterparts leading to better conversion and accounting.- Mapping Amazon products to external sources such as Wikipedia, IMDB and others and extracting relevant information such as book, movie and music reviews and awards to augment the data in the amazon catalog (with attribution).

Started and managed core content systems teams for web crawling, document processing/page rank calculation and search index building in India for Yahoo Web Search. Also, led the engineering team which worked closely with Yahoo research on productizing key technologies in semantic information extraction for its use in Web Search and other areas in Yahoo. The team was responsible for developing highly parallel applications running on Hadoop and Yahoo internal clusters consisting of thousands of machines.

Managed a team working on an end to end solution for dectecting and reacting to security threats in the network using approximate matching techniques like bloom filters, elephant traps and approximate counters. The team was responsible for on chip firmware, device drivers, control plane software as well as the management application for configuring the feature and managing the alerts. Additionally, functionally responsible for QA resources for testing this feature.Filed 2 patents for enhancements to bloom filters.

Worked on the MDS line of storage area networking routers. Part of the infrastructure and platforms team. Strong contributor to the control plane software design. Responsible for application level high availability, port bringup and finite state machine execution engine. Developed (and filed patents for) exchange peer protocol (EPP) for link level parameter exchange.

Worked in the area of embedded system design automation. Proposed a framework for automated partitioning of a system to hardware and software portions based on identification of bottlenecks given a set of constraints.

Worked on asic design flow automation.