Role Overview
DoorDash's GenAI Platform team builds shared infrastructure for GenAI-powered products across DoorDash, Wolt, and Deliveroo. The team runs frontier open-weight LLMs/VLMs (GLM, Qwen, Kimi, DeepSeek) with real-time GPU serving, batch inference, and fine-tuning, achieving large cost/latency wins. They own the LLM Gateway, Agent Gateway, evals infrastructure, guardrails, and cost attribution.
Responsibilities
- Build infrastructure to move GenAI ideas from prototype to production, increasing AI business impact velocity.
- Work on open-weights serving stack: real-time GPU endpoints, high-throughput batch inference, fine-tuning (SFT/DPO/LoRA), LLM Gateway, Agent Gateway, evals, guardrails, cost attribution.
- Design scalable, high-performance systems for model serving, batch inference, GPU autoscaling, and fine-tuning.
- Push cost/latency frontier of GPU inference, turning batch jobs from days to hours and cutting inference cost.
- Build platforms supporting rapid experimentation with production standards (latency, scale, monitoring, SLOs, playbooks).
- Partner with ML engineers, product engineers, data scientists, and platform teams across DoorDash, Wolt, Deliveroo.
- Shape future of centralized GenAI platform including RLHF/RLVR, agent optimization, post-training and agentic techniques.
Requirements
- B.S., M.S., or PhD in Computer Science or equivalent
- 3+ years industry experience in software engineering
- Strong backend engineering fundamentals, especially Python and distributed systems
- Experience building production services, APIs, data pipelines, or ML infrastructure at scale
- Experience operating systems in production: observability, debugging, reliability, incident response, performance/cost optimization
- Hands-on experience with LLM inference and/or fine-tuning of open-weight models in production (serving: latency, throughput, batching, autoscaling, GPU utilization; fine-tuning: SFT/DPO/LoRA)
- Ability to work across ambiguous, fast-moving technical areas and turn customer use cases into reusable platform capabilities
- Proficiency using AI coding tools (Claude Code, Codex, Cursor) in full SDLC
Nice To Haves
- Experience with LLM inference engines and serving frameworks (vLLM, SGLang, TensorRT-LLM) in production
- Experience with distributed/multi-node training and fine-tuning
- Experience with GPU cluster management and scheduling (Slurm, Kubernetes)
- Experience with ML observability and evaluation frameworks
- Contributions to open-source ML infrastructure projects