Engineering

AI Engineer

Full-timeRemoteMultiple openingsReports to: Founder and Chief Executive Officer

Build the AI agents that find and fix what costs small businesses customers, with the owner approving every change.

About SpareBrains

SpareBrains helps small and medium-sized businesses run better. A team of people and AI agents does real work for them. It finds where they’re losing customers online, fixes it, and keeps it running week after week. Our agents work inside our clients’ own accounts and tools, always with the owner’s approval. That makes trust the heart of what we sell.

The role

You’ll build the AI agents at the heart of SpareBrains. They check a business’s listings, calls, reviews and bookings, draft the fix, and make it once the owner says yes. You’ll own them end to end, from the first prototype to running in production every day. You’ll decide how they work, which tools they use, and how we know they got it right.

What you’ll do

Agents that do real work

Design and ship agents that check a business’s online presence and fix what’s wrong, inside the owner’s own accounts.
Design each agent’s tools, prompts and context, and choose the right model for each step.
Turn what our people do by hand into dependable, repeatable agent workflows.

Evals and quality

Build evals from real cases, with golden sets, regression suites and human review, and let them decide what ships.
Measure accuracy, cost and speed, and improve all three over time.

Safety and approvals

Build the approval step: agents propose, the owner says yes by text, and a person checks every item.
Give agents scoped, time-limited access through named logins, and log every action they take.
Guard against prompt injection from the web pages, reviews and messages agents read.

Running in production

Own orchestration, state, retries and recovery for agents that keep working week after week.
Add tracing and monitoring, so we can see why an agent did what it did.
Share on-call for the agents you ship.

Example projects

An agent that finds where a business’s hours, phone and links disagree across Google, Yelp and its own site.
A missed-call follow-up that texts back from the owner’s own number, and knows when to hand off to a person.
Review replies drafted in the owner’s voice, waiting for their OK.

Shaping the company

Work with our forward deployed engineers and client team, and turn their lessons into product.
Help choose our stack, our standards and the next engineers we hire.

What you’ll bring

4+ years building production software, with at least one year shipping features built on large language models.
Strong Python or TypeScript, and good judgement about APIs, data and systems.
Hands-on experience with tool-using agents, evals, and prompt and context design.
You care whether the agent got it right for a real person, not only whether the demo worked.
Clear writing, and comfort turning a vague problem into a plan.
You use AI coding tools every day, and know where they help and where they don’t.

You don’t need every item on this list. If the work excites you, apply anyway.

Nice to have

Experience as a founder or a founding engineer.
Browser automation, or the Model Context Protocol (MCP).
Voice, telephony or SMS systems.
Durable workflow engines, queues and job scheduling.
Work with small businesses, or in local marketing, hospitality or the trades.

What we offer

A competitive base salary, equity, and the chance to build SpareBrains from its first day.

How to apply

Send your name, your email and your resume, with links to your work if you like. No cover letter needed.

Apply

Interview process

Intro call. 30 minutes with our founder on the role, your work and your questions.
Role interview. A deeper conversation about how you’d do the job, and the work you’ve done.
Working session. We work through a real problem together, with any tools you like, AI included.
References and offer. We speak to people you’ve worked with, then send the offer in writing.