Vikas Kedigehalli
Senior Staff ML Engineer Tech Lead Manager at Google
California, United States
Summary
👤
Senior🎓
Top SchoolVikas Kedigehalli is a Senior Staff ML Engineer and tech lead manager with 13 years of experience building large-scale recommendation and ads systems at Google, where he currently drives post-training efforts for YouTube Core Recommendations and Gemini-based conversational recs. He has deep expertise in retrieval, ranking, auctioning, and production ML pipelines from work across Discovery Ads, Retail Ads, and the Cloud Dataflow/Apache Beam stack. A hands-on engineer who still contributes to open-source data-processing tooling, he has enhanced Apache Beam and Cloud Dataflow DatastoreIO components that power batch and streaming pipelines. Based in California with an MS in Computer Engineering from Georgia Tech, he combines systems-level engineering with product-focused ML deployment, often surfacing non-obvious improvements in data handling and testing that improve model quality and operational reliability.
13 years of coding experience
5 years of employment as a software developer
BE, Electronics & Communications, BE, Electronics & Communications at R. V. College of Engineering, BangaloreMS, Computer Engineering, MS, Computer Engineering at Georgia Institute of Technology
Stackoverflow
306reputation
4kreached
8answers
0questions
Github Skills (15)
data-processing10
javas10
google-cloud-dataflow10
sdk10
google-cloud-datastore10
apache-beam10
java10
testing10
api-design9
streaming9
big-data9
google-cloud-platform6
google-app-engine6
python6
mysql6
Programming languages (2)
Github contributions (5)
apache/beam
Jun 2016 - Sep 2017
Apache Beam is a unified programming model for Batch and Streaming data processing.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:108 commits, 129 PRs, 553 comments in 1 year 2 months
Contributions summary:Vikas's contributions are centered around the Apache Beam project, primarily involving improvements to the Java 8 examples, specifically in the context of the LeaderBoard test case. The commits demonstrate the implementation of various testing scenarios within the LeaderBoardTest class, including tests for handling on-time, speculative, and late data within the specified windows and allowed lateness configurations. The user is involved in enhancing the functionality of the test to better validate the team scores.
golangpythonstreaming-databeambatch
GoogleCloudPlatform/DataflowJavaSDK
Jul 2016 - Dec 2016
Google Cloud Dataflow provides a simple, powerful model for building both batch and streaming parallel data processing pipelines.
Role in this project:Back-end Developer
Back-end DeveloperContributions:11 commits, 13 PRs, 21 comments in 5 months
Contributions summary:Vikas contributed to the Cloud Dataflow Java SDK, focusing on the DatastoreIO module. They added support for v1beta3 versions of Datastore Source/Sink, including new classes and deprecating older ones. The user also added functionality for deleting entities and keys, writing mutations, and incorporated improvements for statistics queries. The user's work focused on extending the SDK's interaction with Google Cloud Datastore.
stream-processingbeambatchdata-processingparallel
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.