Founding Engineer — AI Systems & Product

What we’re building

We’re building a system that helps ecommerce companies improve their websites dramatically faster.

The product takes a site, creates a safe sandbox, generates conversion experiments, and sends AI-powered synthetic shoppers through each version. It then helps determine which changes are actually worth testing with real customers.

That sounds simple. It is not.

There are browser agents, behavioral simulation, distributed execution, analytics integrations, experiment design, statistics, frontend product work, infrastructure, and a healthy amount of “why is this browser doing that?”

On a random Tuesday, you might:

• Debug state leakage across hundreds of browser agents.
• Figure out why synthetic shoppers behave differently from real traffic.
• Build an interface for reviewing 100 proposed experiments.
• Improve an experiment runner from 20 minutes to 4.
• Decide whether an apparent conversion lift is real or nonsense.
• Ship the whole thing so a nontechnical customer can use it.

We’re looking for someone who finds that list exciting rather than alarming.

Who we’re looking for

We are looking for a deeply technical, highly curious generalist.

You do not need to know our exact stack. You do need to be the kind of person who can enter an unfamiliar codebase, form a mental model quickly, and start making useful changes without waiting for a twelve-page specification.

You probably:

• Go down technical rabbit holes because you genuinely need to understand how something works.
• Have built things that nobody asked you to build.
• Can move comfortably between product, backend, infrastructure, data, and frontend problems.
• Care about correctness, but also know when something is good enough to ship.
• Have strong opinions, explain them clearly, and update them when the evidence changes.
• Notice when something is technically functional but confusing, fragile, or untrustworthy.
• Use AI tools aggressively without outsourcing your judgment to them.
• Are low-ego, direct, dependable, and unusually good at figuring things out.

We care much more about your slope than your pedigree.

What you’ll own

You will help design and build the core product end to end.

That will likely include:

• Building systems that run large numbers of browser-using agents reliably and in parallel.
• Creating sandboxed versions of ecommerce sites and safely applying UI variants.
• Developing synthetic-user calibration and experiment orchestration pipelines.
• Building statistical and behavioral guardrails around experiment results.
• Integrating with analytics platforms, ecommerce platforms, and live testing tools.
• Creating product interfaces for configuring experiments, reviewing recommendations, and understanding results.
• Improving system speed, reproducibility, observability, and cost.
• Working directly with the founders and early customers to turn messy problems into useful product.
• Helping establish the engineering culture, architecture, and standards of the company.

You will have a lot of ownership. That means freedom, but it also means there may not already be a person, process, or document telling you exactly what to do.

Strong signals

Any of these will get our attention:

• You can show us a technically interesting product, tool, library, experiment, or side project you built.
• You have owned a difficult system in production and can explain what broke.
• You have taught yourself a new technical domain quickly enough to ship something real.
• You have experience working across the stack rather than staying inside one narrow layer.
• You enjoy benchmarking, instrumenting, reverse-engineering, profiling, or reading source code.
• You can explain both how a system works and where its conclusions might be wrong.
• You have a story about changing your mind after the data contradicted your original theory.

Useful experience, but not a checklist

Experience with any of the following is helpful:

• Browser automation, Playwright, Puppeteer, or computer-use agents.
• LLM applications, agent orchestration, evaluation, or synthetic data.
• Distributed systems, queues, containers, concurrency, or job scheduling.
• Experimentation platforms, statistics, causal inference, or A/B testing.
• Product analytics, Google Analytics, Shopify, or ecommerce infrastructure.
• Full-stack product engineering.
• Performance optimization and production observability.

Nobody will have all of this. That is fine.

Things we do not care much about

• Whether you went to a particular school.
• Whether you worked at a famous company.
• Whether you have exactly eight years of experience.
• Whether you already know every technology in our stack.
• Whether you can solve an obscure algorithm problem on a whiteboard from memory.
• Whether your résumé is beautifully formatted.

We care whether you can learn, build, debug, reason, communicate, and finish.

This probably is not a fit if

• You need every requirement to be fully specified before starting.
• You want to stay inside one narrow technical lane.
• You enjoy architecture discussions more than shipping working systems.
• You treat frontend quality, reliability, instrumentation, or customer experience as someone else’s job.
• You confidently repeat AI output without checking whether it is true.
• You consistently reach for the most complicated solution available.
• You want the structure and specialization of a large engineering organization.

Why join

• You will work on a genuinely difficult technical product with unanswered questions.
• You will influence the architecture and product rather than inheriting them.
• You will work directly with customers and see whether what you built actually matters.
• You will have meaningful ownership and equity.
• You will ship quickly without pretending that reliability and rigor do not matter.
• You will work with people who care more about solving the problem than performing startup theater.