Quality Engineer (AI-Native QA)

About Alva

Alva is your AI investing agent. Research a thesis. Monitor a narrative. Backtest an idea. Automate a strategy. Alva turns any of it into a live Playbook — with a team of agents watching the market 24/7 and alerting you the moment it matters. Powered by institutional-grade data from 100+ sources and a growing library of community skills and playbooks you can follow, remix, and learn from, Alva is your personal edge in all markets.

Alva was founded by a team of proven serial entrepreneurs. The founding team includes co-founders from Galxe, a major Web3 growth platform with over 30 million active users, as well as a co-founder from MiniMax, a leading frontier AI lab.

About the role

Most QA job posts in 2026 still read like 2018: write the cases, maintain the regression suite, file the bugs. This is the opposite. Writing cases, running exploratory passes, patrolling production — most of that work goes to agents. You set the strategy and hold the gate.

You own product-feature quality: test strategy, the regression system, and automation — built from scratch and kept running.

What you'll do

• Quality process — test strategy, regression checklists and bug flow across three surfaces: Web, Mobile App, IM Bot
• AI-native automation — an automation base (Playwright + CI) with agents wired into daily testing: cases generated from specs, exploratory passes on PR previews that come back with screenshots and recordings, auto-triage of failures
• Production agent patrols — scheduled journeys walked as a real user (sign up → create an Automation → build a Playbook → receive an Alert), checking availability, alert delivery and financial-data correctness. If a daily change is computed off the previous close instead of the adjusted price, the patrol should catch it — before any user does
• Metrics and the release gate — define regression pass rate and escape rate so “ship or not” is an evidence-based call
• Bad-case loop — trace quality issues from user feedback and monitoring, drive the fix, and turn it into a regression case or eval case. Agent output quality belongs to the eval system; you own the last mile on the product side

What we're looking for

• 3+ years in software testing with strong fundamentals: case design, edge cases, and bug reports an engineer reads once and understands
• Fluent in one automation stack (Playwright / Cypress + TypeScript, or Python), including building a framework from scratch and wiring it into CI
• AI First — Claude Code, Cursor, Playwright MCP in your daily workflow, with concrete examples of how AI changed the way you test. LLM / eval experience is a bonus, not a requirement
• Genuine interest in investing — you read charts, filings and the usual metrics; ideally you've traded yourself. You're testing an investing product, so how much you understand decides how deep your bugs go
• Hands-on and evidence-driven: never satisfied with “looks fine” — you reproduce it yourself and argue with data
• Founder mindset: you build a process where none exists, and you want the bandwidth of working onsite
• Bonus: testing AI / agent products, RAG or agent traces; eval tooling (Promptfoo, DeepEval, LangSmith); quality work in fintech, brokerage, trading, or market-data products; synthetic monitoring, CI release gates, API or data-quality testing; a public repo, tool or write-up

Why this seat is rare

• Most “AI QA” roles don't really exist yet — QA postings at frontier AI companies are still Playwright + Datadog, with no LLM or agent testing in sight. Here the method is deliberately undefined: you decide what it looks like
• AI × finance crossover — eval design, agent testing and financial-data quality are scarce experience anywhere
• You build the quality system 0 → 1 and grow into the quality lead as the team scales
• The take-home is honest work, not a brainteaser: half a day to a day actually using Alva, finding real bugs, then automating one journey — with AI doing most of the typing and you telling us which of its output you overrode, and why

Compensation

¥350K+ base + bonus + equity. Full-time. Location: Shanghai · Onsite.

How to apply

Email [email protected] with something you've built and a few lines on why this role.

Full details: https://alva-careers.netlify.app/playbook