Description
We're looking for a scrappy, resourceful Developer Advocate who understands inference to help developers run open models in production on Token Factory, Nebius' high-performance inference platform.
Your job is to have credible, hands-on answers to questions developers have when choosing an inference layer, and to show the work in public. You'll be embedded with the Token Factory product and engineering team in San Francisco, participating in standups and bug bashes, carrying customer stories back to them, and fixing documentation gaps.
Responsibilities:
- Help developers and teams discover, evaluate, and adopt Token Factory for production inference workloads.
- Build demos, benchmarks, and reference architectures that make real tradeoffs concrete.
- Show how Token Factory fits into the stack developers already use.
- Own the developer feedback loop, gathering feedback and bringing it back as concrete product input.
- Represent Nebius in the Bay Area inference and open-model community.
- Publish technical content and give talks that take developers from first hearing about Token Factory to running something on it.
Requirements:
- Hands-on experience running open models in production or at scale through an inference provider or a serving stack.
- Working understanding of what happens under the hood of a serving engine.
- Current, first-hand knowledge of the open model ecosystem.
- Comfort writing and reviewing code, reproducing issues, and shipping fixes to docs and samples.
- A track record of building and explaining in public.
- Ability to explain complex systems clearly to technical and non-technical audiences.
Nice to have:
- Hands-on experience with supervised fine-tuning, distillation, or RL post-training of open models.
- Contributions to open-source inference or distributed-compute projects.
- Experience at an inference or GPU cloud provider.
- An existing audience that trusts your takes on models and infrastructure.