Jibin Parayil Thomas Email & Phone Number

South San Francisco, California, Us

Document DB TeamDevelop tier-0 database infrastructure based on MongoDB to support next phase of Stripe’s growth

Mountain View, California, Us

Data Platform Team* Worked on query optimizers for proprietary in-memory database optimized for analytical queries* Coordinated and led initiatives to meet query performance SLAs which was critical to launch high impact use cases at large scale customers such as Canadian Tire, Verizon, etc* Pitched and developed micro-service to ingest customer logs in near real time which helped immensely with query performance debugging* Developed query generation optimizers to reduce query execution latency and cost* Leveraged Snowflake UDF to implement Customer Journey Analysis feature in ThoughtSpot* Improved data ingestion performance using in-house distributed tracing platform* Filed multiple patents based on query generation/execution which is currently under review

Mountain View, California, Us

Infrastructure Team* Developed infra framework to catch performance and correctness regressions* Reduced memory footprint of in-house Docker based clusters by 40%* Significantly improved ingestion and performance of in-house distributed tracing platform (Similar to Jaegar tracing)* Scaled up infrastructure to ingest structured traces collected periodically for log analysis and debugging which helped in significantly improving ThoughtSpot’s proprietary database performance* Developed framework for query execution performance benchmarking

Minneapolis And St. Paul, Minnesota, Us

Mountain View, California, Us

• Developed tool to replay customer search logs with data-less customer backup for performance bench-marking as well as to reproduce customer search errors in-house for easier debugging• Parsed search logs stored in Google Protobuf format and mimicked search index in order to reproduce auto-complete suggestions specified in the logs while replaying logged searches with data-less customer backup• Set up Hive pipeline for analysis of logs stored in Amazon S3 buckets• Wrote custom Hive SerDe to parse search logs and create required Hive tables• Packaged all the components associated with the Hive pipeline into a dockerimage for ease of ad-hoc deployment

• Worked as Animation Technical Director for the DreamWorks movie-Trolls. Work involved enhancing proprietary animation software and render using Python, debugging technical issues faced during movie production as well as enhancing the movie production pipeline.• Automated several workflows within DreamWorks pipeline using Python scripting which saved thousands of man hours.• Added new features as well as enhanced the performance of the in-house production tracking website

• Worked as Pipeline Technical Director for the DreamWorks movie - The Boss Baby during the first half of the internship. Work involved development of an application using PyQt and studio APIs to automate the transfer of required data to the cross-site studio as well as to validate the integrity of the transferred data in order to ensure that the cross-site artists have the correct data.• During the second half of the internship, worked as Animation Technical Director for Trolls and Motiongate ride in Dubai.

Master'S Degree, Computer Science

M.Sc. (Hons) Mathematics And B.E (Hons) Computer Science, Computer Science