Staff Software Engineer

Staff Software Engineer - Virtual Machine

Azul Systems

Location:
Czech Republic , Prague

Category:
IT - Software Development

Contract Type:
Not provided

Salary:

Not provided

Save Job

Apply Position

Job Description:

As a Staff Software Engineer, you will be a technical leader who designs, implements, and optimizes core VM subsystems—runtime, JIT/AOT compilation, memory management, scheduling, and OS/hardware interfaces. You will set architecture direction, mentor engineers, and collaborate cross‑functionally to ship production‑grade features that move key performance and latency metrics.

Job Responsibility:

Lead the design and architecture of core VM components (runtime services, JIT/AOT compiler pipelines, GC, thread scheduling, synchronization primitives)
Own end‑to‑end delivery of complex features from ideation and design reviews to implementation and benchmarking
Profile and optimize latency‑critical paths across runtime, compiler, OS, and hardware boundaries (NUMA, caches, micro‑architecture, syscall/IO paths)
Establish technical standards and code quality practices
drive design documents, RFCs, and engineering reviews
Collaborate with performance engineering to define and hit SLOs for tail latency, throughput, footprint, and startup/warm‑up
Advance observability in the VM (metrics, tracing, perf counters) to diagnose GC pauses, compilation stalls, safepoints, and contention
Mentor and coach engineers
cultivate a culture of rigorous design, empirical validation, and respectful collaboration
Contribute to roadmap planning, risk management, and release criteria
communicate trade‑offs to stakeholders

Requirements:

10+ years of systems‑level software development with deep expertise in C and C++ (performance, memory safety, concurrency)
Strong knowledge of virtual machine internals: runtime services, class/method metadata, bytecode/IR pipelines, safepoints, deoptimization, inline caches
Familiarity with compilers (JIT and/or AOT): IR design, code generation, register allocation, inlining, tiering, and optimization passes
Solid understanding of operating systems (Linux preferred): processes/threads, scheduling, memory management (NUMA, huge pages), filesystems, networking
Working knowledge of computer architecture and hardware performance: caches, TLB, branch prediction, SIMD, PMU/perf tooling
Proven ability to design clean, scalable architectures and write high‑quality design docs
expertise in API/ABI boundaries and componentization
Comfort with performance analysis tools: perf, eBPF, VTune, flame graphs, micro‑benchmarks
methodical approach to measurement and regression control
Excellent problem‑solving and debugging skills for complex, concurrent systems
Clear communication, cross‑team collaboration, and ability to lead technical initiatives

Nice to have:

Experience with managed runtimes (e.g., JVM, CLR, LuaJIT, WebAssembly engines) or building custom VMs
Knowledge of modern GC strategies (region/mark‑copy, concurrent, pauseless techniques) and memory allocators
Background in low‑latency systems for trading, telco, gaming, or real‑time analytics
Experience with tiered compilation, profile‑guided optimization (PGO), AOT caches, and startup/warm‑up optimization
Exposure to kernel development, drivers, or firmware
familiarity with RT kernels or scheduler tuning
Contributions to open‑source compiler/runtime projects or published performance work

What we offer:

Equity Program - be part of the company success
Annual bonus based on company performance
Referral Program - bring your colleague we would all love to work with
IT Equipment - MacBook Pro or any other HW according to your preferences
Work-life balance - 5 weeks of holidays, 5 sick days, flexible working hours, 100% work from home also possible
Offices in Prague City Centre - if you prefer
Most importantly you will work with top experts worldwide who contribute to the Java ecosystem

Additional Information:

Job Posted:
December 06, 2025

Employment Type:

Fulltime

Work Type:

Hybrid work

Azul Systems - All Job Offers

Job Link Share: