Coral Bricks Ai

Developer Relations San Francisco or remote · Full-time

Make Coral Bricks the obvious default for the people building agent harnesses — through integrations, benchmarks, and workshops

About Coral Bricks

Our mission is to make frontier intelligence affordable and accessible to everyone. Frontier models are finally here — but almost nobody can afford to use them freely. People have token anxiety: they meter every call, ration every context window, and settle for weaker models because the best ones are priced out of everyday use.

We're building the inference platform that ends that, starting with the workloads that feel the squeeze hardest: research and coding agents that swarm across multiple models, plan, call tools for hours, and reason over big context. Classic LLM serving was never built for them — rate limits that throttle real workloads, queues that stretch a 20-minute job into a 4-hour one, costs that grow with every agent turn. Same models, same prompts — many times the tokens per second at a fraction of the cost.

The team is small, technical, and shipping.

The role

Most of the developers we want are already inside an agent harness — OpenCode, Cline, Continue, Goose, Aider, Zed, or one they wrote themselves. They pick an inference provider once, in a config file, and rarely think about it again. This role is about being there when they make that choice, and giving them a real reason to pick us.

It's one loop, not three jobs: ship the Coral provider integration into a harness, publish the benchmark that came out of doing it, then run the workshop where people use it live. Every artifact feeds the next one.

This is a founding go-to-market role, and it is an engineering role. You'll write code that gets reviewed by maintainers who don't work here, and numbers you publish will be re-run by skeptical readers. Both need to hold up.

What you'll work on

• Ecosystem integrations. Get Coral Bricks shipped as a first-class provider in the open-source agent harnesses developers already use. That means real PRs into other people's repos: provider adapters, docs, config examples, and the follow-through to keep them working as those projects move.
• Technical content with original numbers. Benchmarks, teardowns, and engineering writeups where the data is yours — you ran it, you can defend it, and a reader can reproduce it. Our audience is CTOs, ML engineers, and Python developers who discount anything that reads like marketing, so this is the whole bar.
• Workshops and events. Design and run hands-on sessions at conferences, hackathons, and developer meetups where people leave with something working against our API — not a slide deck they sat through.
• Partner follow-through. We'll open the door with harness maintainers and partner teams; you own everything after the handshake — the integration, the docs, co-marketing, and the relationship as it compounds.
• Bringing the feedback back. You'll be the first to hear where onboarding breaks, which model in the catalog people actually want, and what the docs don't say. That belongs in front of the engineering team while it's still fresh, not in a quarterly summary.

You probably have

• Enough engineering depth to work unsupervised in an unfamiliar codebase: comfortable in Python and TypeScript, comfortable opening a PR against a project you didn't write.
• Real familiarity with LLM inference and the agent tooling ecosystem. You've run models, wired up an agent harness, and have opinions about why some of them feel fast and some don't.
• The ability to generate your own data. You can stand up an inference server, drive load at it, read a throughput curve, and tell the difference between a real result and a benchmarking artifact.
• Writing that engineers finish reading. Direct, first-person, specific numbers, no drama.
• Comfort in front of a room, and comfort traveling for it. Conference presence is lumpy — some weeks are heavy.
• A high work ethic and excitement about early-stage startups. You'll set your own target list and work it without much management.

Bonus

• You've done developer relations or partner engineering at a developer-infrastructure or ML company before, and can point at what came of it.
• Merged contributions to an agent framework, coding harness, or inference project.
• An audience you've built — a blog, a talk series, a maintainer role, a community you're known in.

How we'll work together

You'll report to the CEO and set the target list of harnesses and events with him. We measure this role on things you can point at: integrations merged and live, developers who signed up and made a real API call because of a channel you own, posts published with data you generated, and workshops where attendees left with working code.

Compensation and shape

$120,000–$180,000 base salary, plus 0.25%–1.5% equity, depending on experience. Cash and equity move together — take less of one and we'll weight the other.

Equity vests over four years with a one-year cliff. Health, dental, and vision coverage, and flexible time off. Full-time, based in San Francisco or remote.