Developer Advocate

present of AI; San Francisco (CA); On-site; USD 120,000 - 180,000; Full time

Job summary

Wafer in San Francisco seeks a role blending hands-on engineering with technical content creation. You will build real projects using LLM APIs, write clean sample code, and publish quickly when new models land. A track record in AI or developer tools helps you demonstrate value to engineers and product teams.

You will publish benchmarks, explain inference speed, and contribute demos that help the wider developer community trust and adopt faster AI capabilities.

Qualifications

• Built real projects with LLM APIs and published results.
• Writes clean, runnable sample code for others to reuse.
• Publishes quickly when new models land and shares results with the community.
• Has a record of building work in AI or developer tools for broader adoption.
• Can explain inference speed in a way engineers trust.

Responsibilities

• Publish code and results for new models with clear methodology.
• Create benchmarks and scripts to evaluate latency and throughput.
• Explain inference speed to skeptical engineers and stakeholders.
• Develop tangible AI/demo projects that showcase capabilities.

Skills

LLM APIs

Code samples

Technical writing

Benchmark scripts

Developer community

Job description

This role is part hands-on engineer, part content producer. It asks for someone who has built real projects with LLM APIs, can write clean sample code, and can publish quickly when a new model lands. A strong candidate also brings a following or a record of building one in AI or developer tools, and can explain inference speed in a way that skeptical engineers accept.

How it fits the AI era

AI helps with first drafts, code samples, benchmark scripts and quick demos, so one person can ship much more in a day. What stays human is judgment about which numbers are fair, honesty when a developer points out a flaw, and the taste to know what other engineers will actually read and trust. The role also sits close to the product itself, which is about making AI models run faster.

How it builds AI experience

Someone in this seat would work daily with open models, LLM APIs and fast inference, and would learn how latency, throughput and quantization shape real apps like coding agents and voice agents. To build this experience beforehand, run the same open model through a few providers, measure the speed with a clear method, and publish the code and results. Building a small voice or coding agent on top of a fast API is a good way to show it.

Worth showing

• Working LLM API projects
• Benchmarks with clear methodology
• Fast, clear technical writing
• Developer community audience
• Inference latency and throughput knowledge

This is our own read, written from the employer's posting and the company's Spotlight. The employer's posting is the source of truth for the role.

About Wafer

An AI inference provider that uses AI agents to tune GPU code, which it says lets open models run faster on AMD and Nvidia chips.