Lead Performance Engineer

Intercontinental Exchange Holdings, Inc.•Atlanta, KS

About The Position

Intercontinental Exchange (ICE), owner of the New York Stock Exchange, is hiring a Lead Performance Engineer for its Clearing Performance Engineering team in Atlanta. The role owns performance strategy for event-driven Java platforms that support critical clearinghouses, where end-of-day processing windows are measured in minutes. The engineer leads in-depth investigations, builds test harnesses, mentors peers, and advises on performance trade-offs across application, JVM, operating system, and hardware layers.

Requirements

  • Bachelor’s degree in Computer Science, Engineering, or equivalent experience.
  • At least 8 years of experience in performance engineering, performance testing, or Java development on high-volume, low-latency systems.
  • Deep JVM expertise, including heap and thread dump analysis, GC log interpretation, and memory and CPU profiling.
  • Java and scripting experience with Python, Groovy, or shell for automation and harness development.
  • Hands-on experience with load generation using JMeter, Gatling, or custom harnesses.
  • Experience with message-based architectures such as Kafka, IBM MQ, or equivalent.
  • Strong written and verbal communication skills.

Nice To Haves

  • Experience building observability data pipelines for metrics collection, processing, and storage.
  • Familiarity with Prometheus, VictoriaMetrics, or Grafana.
  • Practical use of AI tools such as Claude, Copilot, or Cursor for performance work, including GC log analysis, query writing, or harness scaffolding.

Responsibilities

  • Lead performance investigations on CPU-bound, multithreaded Java systems, from application code through JVM, operating system, and hardware.
  • Model system scaling with data volume and instrument counts, and produce capacity projections to guide architecture decisions.
  • Identify time-critical paths in batch and event pipelines, and propose and validate fixes such as head-start or out-of-band processing.
  • Build test harnesses and realistic workloads that catch regressions between versions.
  • Tune thread pools, garbage collection, and heap settings for high-throughput, latency-sensitive pipelines.
  • Turn findings into clear, data-backed recommendations for architects, developers, and business stakeholders.
  • Mentor peers and provide guidance on performance trade-offs.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service