Sign in

fault tolerance, database, Kafka, distributed systems, streaming

Cupertino, CA
February 28, 2020

Contact this candidate


Richard Yu

Contact email:


●Actively contributing to Kafka for two years.

Knowledgeable in Kafka’s internal design, particularly its client-side implementation as well as Kafka Streams

Used Java and Scala in my pull requests.

●Had written a fault injection Jepsen test (fault tolerance) for a database built on the RAFT consensus algorithm.

●Test code here:

●A motivated fast learner in STEM related areas

An example is having learned Clojure in a week


●Solid grasp over C++, Java, and Scala

●Can use Linux shell with reasonable proficiency

●Complete grasp of pre-college mathematics and concepts

●Cumulative GPA: 4.0 (non-weighted)


●Some contributions for Kafka listed below.

Largest contributions being KIP-205, KIP-266

Minor KIPs include KIP-534 (and a tentative one: KIP-408)

KIP-205 involves the user querying for a specific range of keys using given time range

Other contributions I have made include modifications of topologies as well as the joining of Kafka data tables.

●Have also got pull requests accepted to Apache Pulsar

●About the Jepsen test suite (used to test fault tolerance) I had written:

Some faults injected included the failure of nodes as well as the disconnection of servers (preventing them from communicating).

The database is called Shiva. I worked on Shiva during a summer internship in 2019 for It relied on a leader-follower paradigm to maintain consistency among replicas. All reads and writes occurs through the leader.

Learned how to setup the database as well as its basic functions.

●Deep Learning Knowledge: I have some knowledge in AI theory, and have implemented some basic Deep Learning operations, such as the ADAM optimizer and feedfoward networks in C++ (without using external libraries).

●Participates actively in competitions (passed USACO Gold).

Contact this candidate