Bengaluru, Karnataka, India
Specialties: Computer Programming, Data Analytics, Distributed Systems
Presto on Qubole: 1. As a part of adding Hive ACID support in Presto, worked on adding read support for original (non-ACID) files of Hive transactional tables. Contributed it to Open Source PrestoSQL as well. Open Source PrestoSQL Blog: https://prestosql.io/blog/2020/09/23/hive-acid-original-files.html 2. Implemented Per Bucket AWS IAM Configuration Feature to enable Presto to use the required S3 bucket IAM configuration dynamically without Presto cluster restart. Blog: https://www.qubole.com/blog/per-bucket-configuration-support-in-presto/ 4. Worked on Dual and Single S3 IAM Role Support in Presto to enable cluster to assume a custom IAM role to access an S3 bucket. 5. Enhanced Insert Over Directory (IOD) feature to publish a new metric ‘number of files’ (number of files written by the IOD query at S3) in queryInfo. Control plane reads it and waits for that many number files to appear at IOD S3 location to prevent eventual consistency issues before showing results on the UI. 6. Added feature to show spot loss warnings with spot interruption time on all the queries scheduled on the node where spot loss will occur. 7. Setup Jenkins jobs for better orchestration of Presto version upgrade process. 8. Responsible for handling customer issues while on-call, active peer code reviews Rubix: Open source S3 file caching framework by Qubole 1. Added socket connection pooling framework to be used for Bookkeeper and Local Data Transfer Server. 2. Added consistent hashing library support in Rubix for better load distribution in case of cluster upscale/downscale. 3. Active OSS contributor to enhance stability and fixing issues. Tech: Java, AWS, Presto, Hive, Python
1. As a part of Cloud Vision Team, working on Authentication Authorization (AA) Service, developed various REST APIs for login, logout, configure and use local, external AA. 2. Worked on Session Management Service which manages sessions for user activities and tasks scheduled by a user. 3. Developed various REST APIs for User and Role Management service to manage cvp user data and provide role based access to a cvp user. 4. Build path based authorization mechanism for Cloud Vision dataset in Aeris(NoSql DB). 5. Wrote several end to end product tests for Python based test infrastructure. Tech: Golang, Java, Hbase, Kafka, Grpc
1. Worked on a new confidential product. I have developed performance utility framework for the graph database TitanDB to benchmark the performance for various graph operations, produce detailed report with different performance metrics and dump the performance data in time series DB(InfluxDB). Platform: Java, Apache Strom, Kafka, TitanDB, Python, InfluxDB, Go 2. As a part of Identity Service Engine(ISE) Development Team, I worked on Monitoring & Troubleshooting Node which is responsible for the logging and reporting functions of ISE and developed various REST APIs. Platform: Java, RESTful Architecture, Hibernate
Developed a “Quantified Posting Management System” to serve as a Decision Support Tool which suggests an ideal candidate for the given post in an organization based on certain rules set by the organization. Tools Used: Drools, Java, Ibator, Ruby on Rails.