Software Development Engineer in Test (SDET)

Software Development Engineer in Test (SDET) — AI-Driven Testing

We operate a regulated digital asset exchange, broker, and custodian serving institutional and professional investors. Our QA practice has been rebuilt around agentic AI: test planning, test code, failure triage, and even code review happen in partnership with AI coding agents, with the engineer as the human in the loop. We're looking for an SDET to join that team. Judgment, curiosity, and clear writing matter more to us than a perfect CV — if you're strong, we're flexible on the rest.

What the role actually is

You won't spend your days hand-writing boilerplate test scripts. You'll direct AI agents to do the mechanical work, and your job is everything the machine can't be trusted with: deciding what's worth testing, checking that a generated test can genuinely fail, spotting when a "passing" suite is lying, and judging whether a red run is a product bug, an environment problem, or a broken test.

What you'll do

• Author and maintain automated UI and API tests (Playwright, TypeScript) covering the

trading, custody, and customer-portal surfaces of the platform.

• Work with agentic AI tools daily — planning tests, generating code, triaging failures,

reviewing PRs — and critically verify everything an agent produces before it ships.

• Triage nightly regression and smoke runs: separate genuine product bugs from

environment noise and test-side defects, and file clear, well-evidenced bug reports that

a developer outside your domain can act on.

• Keep the suite honest: link failing tests to known bugs, release them when fixes land,

and never let a silent skip or a warning-instead-of-failure hide a real defect.

• Run and maintain suites in CI (GitHub Actions) and help keep the pipelines green, fast,

and trustworthy.

• Keep test cases traceable in our test-case-management tooling.
• Occasionally cross the test boundary and fix what you find: QA works in the same

monorepo as the product, so small bug fixes and improvements to the application itself

are yours to ship, through the same PR and review process the developers use.

• Review teammates' test PRs — and respond to review findings on your own, including

findings raised by AI reviewers — with evidence, not opinion.

• Collaborate with developers, QA engineers, and product owners across a globally

distributed team.

What we're looking for

• Hands-on experience in software testing, QA automation, or software development — typically 3–6 years, though we're flexible for the right person.
• Working knowledge of JavaScript/TypeScript, or clear evidence you can pick a language up quickly.
• A solid grasp of how web applications and HTTP APIs work: requests and responses,

status codes, JSON, authentication.

• Genuine interest in AI-assisted engineering. You should be comfortable making AI tools do real work — and sceptical enough to check the output. If you already use AI assistants for real tasks, we'll want to hear how.
• Excellent written English. In an agentic workflow, writing is the work: prompts, bug reports, test plans, and review comments all succeed or fail on clarity.
• Attention to detail and honest reporting. A failing test you explain truthfully is worth more to us than a green one you can't.

Strong candidates usually bring several of the following — you don't need them all:

• Hands-on experience with Playwright (or Cypress/Selenium).
• Experience building or maintaining a shared test framework — fixtures, page objects, helpers, CI wiring — rather than only writing specs inside one.
• Exposure to financial services, trading, or capital markets — order lifecycles, settlement, custody concepts.
• WebSockets / real-time data, SQL/PostgreSQL, CI pipelines (GitHub Actions or similar).
• Familiarity with digital assets or blockchain.

Nice to have

• Performance or load testing (k6, JMeter, or similar).

Work Location: Remote