Inference Engineer (AI)

Location: United Kingdom. Candidates must already have the legal right to work in the UK. Visa sponsorship is not available; future relocation to the San Francisco Bay Area may include U.S. visa and relocation support subject to business needs and work authorization requirements.

Base salary: £140,000–£200,000 per year, plus equity and benefits.

Company

hirify.global is an AI research lab developing realtime voice models, inference infrastructure, APIs, and products for consumer-facing applications.

What you will do

• Optimize realtime model serving and sub-second multimodal inference at scale.
• Apply quantization, distillation, caching, continuous batching, paged attention, and speculative decoding.
• Build high-performance inference systems using C++, CUDA, Rust, or highly optimized Python.
• Scale multi-GPU and multi-node inference with Kubernetes, Ray, custom load balancing, and reliable production infrastructure.
• Take models from research through containerization, serving optimization, deployment, and production reliability.
• Profile systems and improve latency, throughput, and stability for thousands of concurrent connections.

Requirements

• Deep understanding of modern serving frameworks and inference optimization techniques, such as vLLM or TRT-LLM.
• Proficiency in C++, CUDA, Rust, or highly optimized Python, with experience profiling NVIDIA GPU workloads.
• Experience with distributed systems, Kubernetes, Ray, load balancing, and multi-GPU or multi-node inference.
• Evidence of building non-trivial systems, contributing to open source, or producing deep technical work.
• Ability to own the full lifecycle from research model to reliable production service.
• PhD in computer science, physics, or mathematics, or equivalent practical experience building backend or ML systems.

Nice to have

• Open-source contributions to major inference engines.
• Deep-dive technical write-ups or other public technical work.

Culture & Benefits

• Work on state-of-the-art realtime voice models used by large consumer-facing AI applications.
• High autonomy in clarifying ambiguous problems through benchmarks and prototypes.
• Performance, latency, reliability, and shipped impact are treated as first-class priorities.
• Flat structure, fast iterations, and minimal process overhead.
• Compensation includes equity and benefits in addition to base salary.