Mobile Manual QA - India

Testmo · Jira · Confluence · Claude Code + the Driivz plugins · Android Studio · Xcode · Charles / Proxyman · Swagger · Firebase · Teams.

AI-era skills we're hiring for (the differentiator)

Fluent with Claude Code as a daily tool. Skills, agents, plugins, MCP servers — you use them to accelerate test design, generate edge cases from a spec, and pressure-test acceptance criteria.

Authoring Testmo cases from a spec with AI scaffolding, then editing with judgement. You know that AI-generated cases are starting points , not deliverables.

You read AI-generated spec analysis and feedback the Testing Guideline section that applies — shaping the downstream test-case generation before it runs.

You spot AI failure modes — hallucinated steps, false coverage, plausible-but-wrong assertions, over-generic cases. You treat AI output like a junior engineer's PR: verify before merge.

You know when NOT to use AI — security-sensitive flows, brand-critical UX moments, exploratory sessions where serendipity is the value, anything with regulatory exposure.

You'll test code that AI developer agents authored and AI PR-reviewer agents approved — you know how these agents fail, which seams they get wrong, and where regression risk concentrates after AI-authored changes.

You communicate in chain vocabulary — gates, contracts, state-transitions, Epics, stories. You think in graded handoffs, not just features.

The bar isn't uses AI tools . The bar is judges AI output well .

How we work

Quality is a gate, not an inspection. We catch issues at PI-week QA scoping, in spec review, and at story decomposition — not in the last sprint of the PI. By the time a story is in-sprint, its Testmo cases already exist.

Risk decides depth. A high-risk Epic gets exploratory + deep regression + cross-flavour. A low-risk one gets the happy paths and the deltas. You make the call and defend it.

You own a feature end-to-end across all three surfaces — Android, iOS, Driver Portal. We don't split testers by platform; the same person follows a feature from spec to ship. Context loss is the enemy.

Pairing over silos. You pair with developers during spec review, exploratory sessions, and defect triage. PMs run their feature templates past you before PI commit. We're async-friendly otherwise.

Claude is a peer, not a replacement. You delegate scaffolding to it, then verify with your own eyes. We share prompts and skill files the way other teams share code snippets.

PIs and sprints. We work in PIs with regular sprints inside. QA scoping happens at the start of every PI — that's where you set the type-mix, environments, and risk rating for each Epic you'll own.

Bonus Points

Espresso, XCTest, KIF, Maestro, Detox, or Playwright familiarity — even just enough to read what the developers wrote.

CI awareness: tests run there, flakiness is a bug, you don't accept "works on my machine" from yourself either.

Performance instrumentation (Android Profiler, Instruments) when the feature has perf risk.

Security and PII discipline — no PII in client logs, ever. (It's a Driivz Mobile rule.)

Prior experience writing or contributing to a team Testing Guideline.

Familiarity with cloud platforms (AWS, GCP, or similar) — useful for understanding the backend side of mobile feature tests.

Knowledge of network protocols (HTTP, TCP) — useful when traces get hairy.

Automation experience that you can bring into manual workflows when it pays off.