Confluent is pioneering a fundamentally new category of data infrastructure focused on data in motion.  Our cloud-native offering is the foundational platform for data in motion --- designed to be the intelligent connective tissue enabling real-time data, from multiple sources, to constantly stream across the organization.  With Confluent, our customers can meet the new business imperative of delivering rich, digital customer experiences and real-time business operations.  Our mission is to help every organization harness data in motion so they can compete and thrive in the modern world.

Dissertation: Discovering Real World Context to Tag Personal PhotosDeveloped a framework called CueNet to discover the context (personal/social/global) of real world events. For example: if a photo was taken from a mobile phone, CueNet could automatically determine if the person was taken at a sports event, dinner party or conference, and tag all the people in it. CueNet combined novel discovery algorithms with data integration techniques and machine learning models (for example, face…
            

    
    

    
        Show more

UC Irvine

Visvesvaraya Technological University

National Public School

Air Force Bal Bharti School, Delhi

Software Engineer

(Turn was acquired by Amobee/Singtel for $310MM)
Worked on Turn's core real time bidding pipeline to facilitate media buying at very large budgets and bid QPS (5.5 MM QPS in Oct 2017).
Designed and implemented Scrub: a real time streaming debugging platform. With Scrub, engineers, account managers and product managers could run real time analytics queries on Turn's vast and complex bid traffic. With this tool, we were able to squash bugs in minutes compared to days and sometimes months, thereby saving the company valuable resources.
Designed and implemented Splicer (https://github.com/turn/splicer), a tool to efficiently distribute and execute TSDB queries.
Developed a CountMinSketch based distributed fraud detection framework (https://patents.justia.com/patent/20180268474).
Designed a real time machine learning pipeline to translate browsing information into actionable segment targeting. What we call pre-targeting (as compared to retargeting).
Developed and maintained the very complex cost distribution models for bid/impression accounting, which determined how much each partner (anywhere from 5 to 30% for each bid) made or paid for each ad impression. This was a very crucial component of the bidding pipeline, where tiny mistakes could be responsible for $100,000 of credits per day.

Platform Engineer

PhD in Computer Science (2013). 
 Thesis topic: "Context Discovery in Real World Events". 
 Doctoral Advisor: Prof Ramesh Jain.
Masters in Computer Science (2010).
Github: https://github.com/wicknicks
Awards: http://www.ics.uci.edu/grad/profiles/index.php?student=satish

Student (M.S. and Ph.D.)

Lives: A timeline based story viewer for historical multimedia data. 
This enabled laying down historical stories right next to each on a large time line. The immediate result was to see how one story line could affect another. For example, when the story of computer engineering and history of communication technology were put together on a single timeline, it became easy to visualize the spike in growth of computer engineering innovations after the construction of the transcontinental telegraph which connected the east and west coast of the United States. Until the laying of this line, the development of technology was stymied. Even with great efforts by people like Charles Babbage the progress was stagnant at best.
Internship Advisor: Gordon Bell.

Research Intern

Software engineer

Create a automated photo indexing system based on the people present in them. Created a SVM based face recognition system for personal photos.