AI SDET

About HubSync:

HubSync is a fast-growing SaaS company revolutionizing how top accounting firms manage their client engagement processes. Our cloud-based platform streamlines workflows and automates complex tasks, empowering firms to deliver exceptional client experiences while increasing efficiency. As we continue to scale, we're looking for passionate team members who thrive in a fast-paced, collaborative environment.

About this role

HubSync is hiring a Senior SDET, Framework & Enablement, to own the health of our end-to-end test automation and make it easy for every engineer to contribute tests that meet the standard. Our customers run their busiest weeks of the year on HubSync, and our test pipeline has to be something they can count on: when a pipeline is red, it means something real; when it is green, the team trusts it enough to ship.

Our test framework is AI-assisted by design. Engineers define what a test should check, Claude Code writes the test code under versioned skill-file rules, and a convention linter plus tiered CI gates (PR, deploy, and nightly) hold the standard. The architecture is settled. This person will own it, extend it, and help the rest of engineering use it well. Success is measured by how many people contribute strong tests, not by how many tests this person writes.

Key Responsibilities

Gate & Signal Ownership

• Triage CI failures to root cause, so every gate failure gets a real disposition instead of a re-run.
• Run test quarantine with discipline: flaky tests go in with a tracking ticket and an owner, and come back out fixed or deleted.
• Lead the quarterly review of the PR gate so only tests that earn their place stay in it.
• Protect signal quality over green builds; a gate that passes because tests were skipped or deleted is a failed gate.

Framework Engineering

• Own and extend the shared test framework: fixtures, factories, the API client, page object conventions, and CI test mapping.
• Lead the move of the convention linter from shell-based checks to AST-aware tooling.
• Build reusable patterns for hard problems, such as complex grid assertions and external signing ceremonies, so feature teams don't have to solve them again.

AI-Assisted Test Development

• Own the skills system that governs AI-written tests; design the rules that define what the model may write and what it must refuse.
• Trace violations in AI-written tests back to their source and fix them in the skill files, not in review comments.
• Re-validate those rules whenever the underlying model changes.
• Review generated tests for convention compliance and real test value.
• Help developers adopt the workflow: pair on their first tests, turn recurring friction into framework fixes or skill updates, and keep the path from test idea to merged, passing test short.

Direct Test Authoring

• Write tests, page objects, and fixtures directly where the stakes or the difficulty call for it: critical-path coverage, reference implementations that set the pattern, and problems too complex to delegate.

Qualifications

• 5+ years in software development in test, test automation, or developer-tooling roles within a SaaS environment.
• Proven experience building or owning a test framework, not just writing tests inside one.
• Hands-on ownership of CI/CD pipelines, including being the person accountable when the build broke.
• Deep TypeScript expertise; Playwright strongly preferred.
• Strong review judgment: the ability to look at a passing test and tell whether it verifies anything.
• Demonstrated track record of making other engineers more productive through internal tooling, developer experience improvements, or documentation teams rely on.
• Genuine interest in AI-assisted development and the governance questions that come with it.
• Experience in tax, accounting, fintech, or another regulated industry is a plus.

What this role is not

• Not a manual QA or test-execution role.
• Not a script factory. Test count is not the goal. Making the whole team able to write good tests is.
• Not a mandate to redesign the framework. The architecture is settled, and this role extends it within established conventions.
• Not the owner of AI feature quality. Evaluating HubSync's own AI features (evals, rubrics, tracing, adversarial testing) is a separate discipline. This role partners on that work but doesn't own it.

Technologies

TypeScript, Playwright, CI/CD with tiered test gates, AST-based linting, Claude Code with versioned skill files, React and Node.js applications, and AWS.

We are an equal opportunity employer. All applicants will be considered for employment without attention to race, color, religion, sex, sexual orientation, gender identity, national origin, veteran or disability status.

Originally posted on Himalayas