29 days ago
London, United KingdomMid Level
Responsibilities
- Build infrastructure that moves GenAI products, agents, automation, and personalization from prototype to production.
- Develop real-time GPU endpoints, high-throughput batch inference, and fine-tuning pipelines using SFT, DPO, and LoRA.
- Design scalable systems for model serving, batch inference, GPU autoscaling, fine-tuning, backend services, and observability.
- Improve the cost, latency, throughput, reliability, fallback behavior, and operational performance of GPU inference and fine-tuning.
- Build platform capabilities with monitoring, SLOs, playbooks, cost controls, and production operational standards.
- Partner with ML engineers, product engineers, data scientists, and platform teams across Deliveroo, DoorDash, and Wolt.
- Help shape future platform capabilities involving RLHF, RLVR, agent optimization, post-training, and agentic techniques.
Requirements
- BSc, MSc, or PhD in Computer Science or equivalent.
- At least 3 years of industry experience in software engineering.
- Strong backend engineering fundamentals, especially in Python and distributed systems.
- Experience building production services, APIs, data pipelines, or ML infrastructure at scale.
- Experience operating production systems, including observability, debugging, reliability, incident response, and performance or cost optimization.
- Hands-on production experience with LLM inference or fine-tuning open-weight models, including serving performance, batching, autoscaling, GPU utilization, SFT, DPO, or LoRA.
- Ability to turn customer use cases into reusable platform capabilities in ambiguous, fast-moving technical areas.
- Proficiency with AI coding tools such as Claude Code, Codex, or Cursor throughout the software development lifecycle.
- Preferred experience with vLLM, SGLang, TensorRT-LLM, distributed fine-tuning, GPU performance optimization, Kubernetes, AWS, GCP, Modal, LLM gateways, developer platforms, AI agents, MCP servers, evaluation systems, observability, RAG, search, or vector databases.
Tech Stack
Categories
About Deliveroo
Deliveroo is an award-winning delivery service founded in 2013 by William Shu and Greg Orlowski. Deliveroo works with approximately 176,000 best-loved restaurants and grocery partners, as well as around 150,000 riders to provide the best food delivery experience in the world. Deliveroo is headquartered in London, with offices around the globe. Deliveroo operates across 9 markets, including Belgium, France, Italy, Ireland, Qatar, Singapore, United Arab Emirates, Kuwait and the United Kingdom.