Shu Zhang
Member Of Technical Staff, Head Of Applied AI at Perplexity
Palo Alto, California, United States
Summary
👤
Senior🎓
Top SchoolShu Zhang is a seasoned engineering leader with a decade of experience building high-performance ML and infra systems, most recently driving AI platform, personalization, and agent products at Perplexity from Palo Alto. As an early member at Pinterest, he helped scale core ads and ML serving infrastructure that supported a 10x valuation and massive revenue growth, growing teams from single digits to 40+ engineers. He combines hands-on systems expertise—contributions to RocksDB and Pinterest’s Rocksplicator replication projects—with product-facing leadership that delivered low-latency, GPU-backed inference and training platforms. Shu mentors teams through large re-architectures, focuses on operational reliability and MLOps tooling, and has a track record of applying database and distributed-systems optimizations to production AI services. Outside work he’s an angel investor supporting early-stage startups, bridging technical strategy with go-to-market momentum.
10 years of coding experience
11 years of employment as a software developer
Bachelor's degree Computer Software Engineering, Bachelor's degree Computer Software Engineering at Tongji UniversityMaster of Science (M.Sc.) Computer Science, Master of Science (M.Sc.) Computer Science at Brown University
Github Skills (14)
rocksdb10
databases10
storage-engine10
c-language10
cprogramming-language10
performance-optimization10
database10
thrift9
apidoc9
pac9
hdfs9
api9
pact9
logging8
Programming languages (4)
Github contributions (5)
pinterest/rocksplicator
Dec 2016 - Mar 2018
RocksDB Replication
Role in this project:
Back-end Developer
Back-end DeveloperContributions:86 commits, 165 PRs, 79 pushes in 1 year 3 months
Contributions summary:Shu primarily focused on developing and refining the `rocksplicator` project, contributing to the core replication functionality. Their work involved implementing a timer for logging slow requests, enhancing the interface and code. The user also addressed comments and improved the overall code quality. The user added functionalities like setting DB options and compacting ranges.
rocksdbreplication
facebook/rocksdb
Dec 2015 - Mar 2017
A library that provides an embeddable, persistent key-value store for fast storage.
Role in this project:Back-end Developer
Back-end DeveloperContributions:12 commits, 6 PRs, 49 comments in 1 year 3 months
Contributions summary:Shu primarily contributed to the RocksDB's back-end functionality by implementing and modifying features related to HDFS environment integration, file ingestion, and FIFO compaction. They added a factory method for creating an HDFS environment and fixed inconsistencies between HDFS and environment configurations. Furthermore, the user enhanced the ingest external file process to improve backward compatibility and enabled a FIFO compaction benchmark, indicating involvement in database performance and optimization.
persistent-storagebigtablefast-storagelsm-treedatabase
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.