Benchmark design & execution: Build and maintain repeatable end-to-end benchmarks for database and AI workloads
Database performance engineering: Evaluate throughput, latency, concurrency, scalability, and stability across OLTP/OLAP and mixed workloads.
AI performance benchmarking: Measure and optimize performance for GPU/CPU inference pipelines, model serving, and data/feature pipelines; assess cost/performance tradeoffs.
Tooling & automation: Develop harnesses, scripts, dashboards, and reproducible environments to run benchmarks consistently.
Analysis & reporting: Produce clear reports, identify bottlenecks, and recommend fixes.
Cross-functional collaboration: Partner with database engineers, AI engineers, the field teams and product management to define performance goals and acceptance criteria.
Performance best practices: Establish standards for test hygiene, statistical rigor, hardware normalization, and benchmark governance.

8+ years in performance engineering, benchmarking, SRE, or systems engineering (database and/or AI systems).
Strong understanding of performance concepts and stochastic modeling: latency percentiles, tail behavior, throughput, concurrency, queueing, saturation, and capacity planning.
Strong understanding of database internals (Postgres or equivalent databases).
Hands-on experience benchmarking databases and/or distributed systems (e.g., Postgres or similar).
Experience with benchmarking tools and harnesses (examples: pgbench, HammerDB, TPC-like workloads, custom harnesses).
Proficient in Python and/or Go for automation and data analysis, strong scripting skills (Bash), skills in eBPF.
Experience with Linux performance profiling (e.g., perf, flamegraphs, strace, iostat, vmstat).
Ability to communicate findings clearly and drive performance improvements with engineering teams.

Principal Performance Engineer, Database & AI Benchmarking

Key skills