Splunk Staff Software Engineer - Performance Optimization & Innovation (PerfOpt)

Cisco; Kraków; On-site; PLN 210,000 - 320,000; Full time

Benefits offered by this job

Competitive compensation

Learning & career growth

Job summary

Cisco in Kraków seeks a Senior Performance Engineer to own end-to-end performance optimization for a large distributed data platform. You will drive hypothesis-driven work, instrument signals, and ship measurable customer impact in memory, I/O, and tail latency.

Apply advanced profiling tools, CI pipelines, and container tech to prevent regressions, while mentoring peers and shaping architecture for scalability and reliability.

Qualifications

• Advanced C++ in production environments, including memory management and move semantics.
• Deep performance engineering experience across CPU/memory profiling and latencies.
• Experience with profiling tools like perf, eBPF, VTune and flame graphs.
• Strong understanding of latency optimization in distributed systems.
• Hands-on knowledge of containers (Docker/Kubernetes) and telemetry pipelines.

Responsibilities

• Lead performance work from hypothesis through instrumentation to measurable customer impact.
• Mentor peers and grow engineers on the team.
• Plan performance work, estimate effort, and lead postmortems to prevent regressions.
• Design petabyte-scale components for memory, concurrency and I/O.
• Instrument production with telemetry and continuous profiling to measure workloads.

Skills

Advanced C++

Performance engineering

Profiling & benchmarks

Latency optimization

System design

Docker & Kubernetes

Python tooling

Linux internals

CI/CD

Go/Rust

Education

Bachelor's degree + 8 years (or Master's +6, PhD +3)

Tools

perf

eBPF

VTune

GitLab CI

Job description

Meet The Team

The Performance Optimization and Innovation (PerfOpt) team improves the Splunk customer experience through deep performance work, and sets Splunk's long-term performance standards through guidelines and design.

We work where performance is won or lost: search and indexing hot paths, cache and bloomfilter efficiency, S3 transfer throughput, I/O workload management, and parallelism in the data retrieval path. It's a C++ core moving petabyte-scale data across a large distributed architecture, and the results land in what customers feel - search latency, ingest throughput, and infrastructure cost. We are also the organization's performance center of gravity: the benchmarking, design patterns, and engineering standards other teams adopt come from here.

Your Impact

Senior performance engineers advise teams on roadmap and architecture, resolve customer performance issues, and set technical direction across partner teams.

Lead performance work from hypothesis through instrumentation and design to measured customer impact, coordinating across teams as needed.

Design petabyte-scale components for memory, concurrency, I/O, and tail latency, guiding architecture for scalability and reliability.

We use profiling, flame graphs, lock contention analysis, benchmarks, and CI regression gates to identify and prevent performance issues.

We instrument production with telemetry, eBPF, and continuous profiling to measure performance in real workloads.

Compare performance options, quantify costs and risks, and present evidence-based recommendations to partners.

Reusable AI workflows shorten performance analysis from days to hours. Engineers set targets, verify results with profiles and benchmarks, and adjust when evidence falls short.

Plan performance work, estimate effort, and lead postmortems to identify root causes and prevent regressions.

Mentor peers and less experienced engineers on the team.

Minimum Qualifications

• Technical depth
• Advanced C++ in production, including memory management , allocators, move semantics , cache-aware structures, and performance trade-offs .
• Deep performance engineering - depth across the below, production experience with several: CPU/memory profiling and flamegraph analysis (perf, eBPF, VTune), including off-CPU and continuous profiling. Lock contention resolution - hot mutexes, false sharing, atomics and memory ordering, lock-free techniques and when not to use them. Latency optimization in distributed architectures - tail latency, queuing, fan-out amplification, backpressure, critical path analysis. Benchmarking - micro and workload benchmarks, sound test design and statistical analysis, sub-system validation CI regression detection. Performance modelling - analytical or capacity models that help predict scaling limits and validate measured results
• Architecture and system design - design and improve existing solutions for performance, scalability and operability, and explain those designs to a critical audience.
• Docker and Kubernetes - performance characterization in containerized environments.
• Strong Python for tooling, benchmark harnesses, telemetry pipelines, and performance data analysis.
• Linux performance internals - scheduler, memory subsystem, page cache, filesystems, block I/O, network stack.
• Git and CI (e.g. GitLab CI) for automating builds, tests, benchmarks, and releases.

Leadership

• Track record of informing technical decisions - quantifying trade-offs, surfacing risk early.
• Distilling sophisticated problems into comparable data points without losing the nuance or hiding uncertainty.
• Clear, concise, high-signal communication across audiences - mechanism-level with engineers, trade-off-level with
• Mentor and grow engineers - coaching junior engineers into ownership, and influencing peers through technical leadership rather than authority.

Experience

Bachelor's plus 8 years of related experience, master's plus 6 years, or PhD plus 3 years.

Preffered Qualifications

• Experience with search or storage engines and large-scale data platforms.
• Splunk internals.
• Hardware-level work - SIMD, NUMA, cache/TLB behavior, PGO/LTO, compiler optimization.
• Go, Rust, or another systems language alongside C++.
• AWS, Azure, or GCP -storage, networking, and instance performance characteristics.
• Terraform, Puppet, or Ansible for reproducible performance test environments.

What We Offer You

• Performance problems at a scale very few companies have - petabytes of data, real customers, and measurable impact from every millisecond you remove!
• A constant stream of new things to learn. We're always expanding into new areas, bringing in open source projects and giving back, and exploring new technologies.
• Exceptionally versatile and ded