Thomas Kielbus
Senior Staff Software Engineer at Cruise
San Francisco, California, United States
Summary
👤
Senior🎓
Top SchoolThomas Kielbus is a Senior Staff Software Engineer in San Francisco with 15+ years building petabyte-scale data pipelines, multi-million-requests-per-second distributed systems, and high-accuracy prediction engines across mapping, routing, autonomous vehicles, adtech, and search. He has led large teams and core platforms at Uber and Cruise and was a founding engineer at rideOS and Red Planet Labs, designing high-performance routing engines, geospatial databases, and worldwide map-processing pipelines that process entire planet graphs in about an hour. Thomas blends deep systems design (custom distributed KV stores, streaming pipelines, sharded graph architectures) with pragmatic cloud-native ops—CI/CD, Kubernetes, gRPC—to deliver low-cost, production-grade infrastructure. Notably, he built HankDB, a compact, petabyte-scale KV engine optimized for massive sequential writes and predictable low-latency reads, and has repeatedly turned research-grade algorithms into globally deployed services.
15 years of coding experience
12 years of employment as a software developer
Master's degree, Computer Science, Master's degree, Computer Science at EPITA: Ecole d'Ingénieurs en Informatique
French, English
Stackoverflow
36reputation
719reached
1answer
0questions
Github Skills (21)
cascading10
key-value10
low-latency10
nosql-database10
high-performance8
synchronize8
gazelle8
trident6
bazel-rules6
java6
unit-testing6
apache-storm6
protobuf5
translation5
rails4
Programming languages (5)
Github contributions (5)
LiveRampArchive/megadesk
Apr 2012 - Jun 2014
Contributions:306 commits in 2 years 2 months
zookeepersynchronizestepson-topjava
LiveRamp/hank
Feb 2011 - Sep 2014
(DEPRECATED. This project is no longer used or maintained at LiveRamp.) Hank is a high performance distributed key-value NoSQL database that we built and use at LiveRamp. It is designed for very large data stores that dwarf the amount of available main memory and for randomly distributed read/write workloads that far exceed the capacity of memory-based caches. More specifically, it is optimized for very low latency random read queries and for very high throughput incremental batch writes.
Contributions:1872 commits in 3 years 6 months
key-valuelow-latencynosql-database
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.