Associate AI Quality Engineer

The job

Test automation

• Build backend API, contract, service-level, integration, and data-driven test coverage.
• Develop web end-to-end, component, visual, and RTL regression coverage with fast merge-gating suites.
• Develop mobile coverage with Appium or Maestro, including device farms, offline and sync behavior, and payment peripherals.
• Maintain shared fixtures, test data, environments, parallelization, and meaningful CI pipelines.

AI quality systems

• Generate tests from specifications, code, and production traffic while addressing long-term maintenance.
• Classify build failures, detect flakes, quarantine unstable tests, and analyze suite health and coverage gaps.
• Build evaluation infrastructure with datasets, scoring, regression detection, and region-specific correctness checks.

Agentic AI and orchestration

• Build agents that read diffs, run relevant suites, reproduce failures, propose fixes, and open pull requests.
• Create end-to-end quality workflows for exploratory testing, coverage-gap analysis, and release-risk assessment.
• Design orchestration with planning, tools, retries, state, memory, sandboxing, multi-agent handoffs, and human escalation.
• Integrate with CI, Jira, observability, and MCP-style tool interfaces.

Technical qualifications

• Production software engineering experience with recent hands-on work on LLM-backed systems used by real users.
• Strong Python and comfort with at least one of .NET, Java, or TypeScript.
• Experience owning release-gating automation across backend and web or mobile surfaces.
• Experience building evaluation systems and explaining improvement with numbers.
• Practical knowledge of prompting, structured output, retrieval, tool use, agentic orchestration, testing, and CI/CD.

Technical ground

• Knowledge of test framework design, test layering, flake economics, parallel execution, mobile and cross-browser testing, and CI/CD gating.
• Knowledge of context engineering, retrieval, chunking, reranking, caching, and long-context behavior.
• Knowledge of offline and online evaluation, LLM-as-judge limitations, human review, statistical significance, and regression gates.
• Knowledge of structured output, guardrails, retries, non-determinism, tracing, prompt and version management, model routing, latency, and cost controls.

Working approach

• Bias toward adoption and measure work by what other engineers use rather than by demonstrations.
• Apply judgment about when a plain pipeline is better than an agent.

Work arrangement

• The posting mentions flexibility in office and hybrid work setups but does not clearly specify the arrangement for this role.

What Foodics offers

• The posting mentions competitive compensation, bonuses, potential shares, training, a learning stipend, autonomy, mentoring, and challenging goals.
• Foodics describes an inclusive culture and a distributed team across multiple countries and nationalities.