Senior Staff Engineer
Los Angeles preferred, remote OK within US time zones · Full-time
Checkout Bench tests whether AI shopping agents can actually buy things. We run agents through real, signed-in store checkouts at Amazon, Walmart, Target, Instacart, Uber Eats and Chewy, and grade every run in code instead of with another model. Across 527 real checkout requests so far, only 35% made it to a ready quote at checkout. The biggest drop is between search results and reading the checkout page. Retailers use us to find where agents get stuck on their sites, and agent builders use us to see where their agents break.
What you'll do
- Own the architecture of the run engine. Every run gets a fresh cloud browser, signs in to a real store and builds a cart through to a ready quote at checkout.
- Make runs reliable and reproducible at scale, with full traces that can be replayed step by step.
- Design the graders in code: right item, right quantity, within budget, and asking when the request is unclear.
- Build the private test environments we stand up on retailers' own catalogs and checkout flows.
- Set the engineering bar, review the hard designs and mentor the team as it grows.
What we're looking for
- You have 10+ years building production systems, including distributed or browser-heavy infrastructure.
- You have deep experience with headless browsers and web automation (Playwright, Puppeteer, CDP) and with the messy reality of real websites.
- You have strong instincts for evaluation, observability and data integrity. You'd rather have a number you can defend than a flattering one.
- You're comfortable working directly with LLM agents and their failure modes.
- You can ship fast on a small team and still leave systems others can build on.
Nice to have
- Experience with e-commerce, payments or checkout flows.
- Work on benchmarks, eval harnesses or RL environments.
Location
We'd love you in Los Angeles to work in person with the team, but remote within US time zones is fine for the right person.
Pay
$200,000–$250,000 base + equity
When you apply, include something you've built that you're proud of.
Forward Deployed Engineer
Los Angeles preferred, remote OK within US time zones · Full-time
People are starting to hand their shopping to AI assistants. When an assistant can't find the right size, gets stuck on a pop-up or stalls at checkout, the sale doesn't happen and nobody tells the store. Checkout Bench finds those failures before customers do. We run AI shopping agents through real, signed-in checkouts and grade every step in code.
As a Forward Deployed Engineer, you're the person who turns that into results for a specific retailer.
What you'll do
- Work directly with retailers' e-commerce and growth teams to build a private test environment on their own catalog, variants and checkout flow.
- Turn the requests their customers actually make into test cases, then run assistants like ChatGPT, Gemini and Perplexity through them.
- Diagnose exactly where agents get stuck (search, product options, sign-in, cart or checkout) and separate the store's problems from the agent's.
- Deliver fix lists ordered by how many runs each problem blocked, then re-check after the retailer ships changes.
- Bring what you learn in the field back into the product and the engine.
What we're looking for
- You have 4+ years as a software engineer, solutions engineer or similar, writing real code in a customer-facing role.
- You're fluent with the web stack and browser automation. You can read a checkout flow and see where an agent will trip.
- You write clearly and communicate calmly with both engineers and business stakeholders.
- You have a bias to action. You'd rather prototype a fix than write a long memo about it.
Nice to have
- Experience at or with retailers, Shopify or other commerce platforms.
- Hands-on experience building or evaluating LLM agents.
Location
Los Angeles is preferred, and we'll be together in person often. Remote within US time zones works too, with some travel to customers.
Pay
$150,000–$200,000 base + up to 2% equity
Director of Content Marketing
Los Angeles preferred, remote OK within US time zones · Full-time
Checkout Bench runs AI shopping agents through real store checkouts and publishes what happens. Our data shows things nobody else can see yet. Of 527 real checkout requests, 186 made it to a ready quote, and most failures happened after the agent had already found the product. Our job is to make retailers, model labs and agent builders care about that, and we need someone to own the story.
What you'll do
- Own content strategy and execution across LinkedIn, X, our site, newsletters and launches.
- Turn benchmark data and run traces into reports, charts, short videos and posts people share and cite.
- Build our founder-led LinkedIn presence and a regular rhythm of findings retailers look forward to.
- Write and edit everything from landing page copy to research write-ups, keeping it plain, precise and honest about what the numbers do and don't say.
- Work with engineering and design on launches, case studies and the public benchmark.
- Hire and manage writers, designers or agencies as we grow.
What we're looking for
- You have 7+ years in content or product marketing, ideally at a technical or data-heavy B2B company.
- You have a portfolio showing you can make technical material clear and compelling without hype.
- You're comfortable with data. You can find the real story in a chart and refuse to overclaim it.
- You have strong instincts for social distribution, especially LinkedIn and X.
- You have high taste in writing, design and motion.
Nice to have
- A background in e-commerce, retail tech or AI.
- Experience launching a benchmark, research report or developer product.
Location
We'd love you in Los Angeles, but remote within US time zones is fine.
Pay
$150,000–$200,000 base + equity
When you apply, send two or three pieces you're proudest of.