Junior AI Quality Engineer at Foodics, Riyadh

Most quality job ads ask for someone who writes tests. Foodics' new posting in Riyadh asks for someone who builds the agents that write and maintain them, then measures whether engineers actually use them. The title is junior, but the task list reveals a role that sits at the intersection of software testing and large language model systems.

Role:Associate AI Quality Engineer at Foodics, in Riyadh, Saudi Arabia, with full-time on-site presence as part of the technology team and the quality assurance department. Foodics is a restaurant-management and payment-technology platform founded in 2014, headquartered in Riyadh, with offices in the UAE, Egypt, Jordan and Kuwait, and its posting states that its products have processed more than 6 billion orders and that its latest funding round raised $170 million.

Primary challenge:Build the internal systems that the company’s engineers use daily: agents that generate and maintain tests, pipelines that classify build failures before a human opens them (real defect, flaky test, environment issue, or obsolete test), and tools that accelerate code review and bug tracking. The role also extends to the architecture for evaluating AI features in the company’s own products: datasets, evaluation and regression detection when a model or prompt changes. The posting is explicit about a point that concerns our market specifically: evaluation must cover system behavior in Arabic and English, right-to-left writing interfaces, and the region’s point-of-sale, tax and payment rules, where the correct answer is rarely a literal text match.

Key requirements:

• Recent hands-on experience with large-language-model-based systems that real users rely on, and production-grade tested code rather than experiment notebooks.
• Proficiency in Python and comfort with at least one of.NET, Java or TypeScript.
• Depth in test automation across multiple interfaces, using tools such as Playwright, Appium and Maestro, and experience owning a test suite that is version-controlled.
• Understanding of agent orchestration, tool usage and standards such as MCP, and context engineering involving retrieval, chunking and re-ordering.
• Familiarity with evaluation methods: direct and indirect assessment, using a language model as a judge and its limitations, and rollback gates within CI.

Salary:Not disclosed.

Practical step:Before applying, prepare a single documented example from your work: a test suite you kept stable without removing difficult tests, or an evaluation system for an AI feature that actually detected regression. The posting states clearly that the metric is engineers’ adoption of what you build, so showcase impact rather than a tool list. If you have experience evaluating Arabic outputs, place that front-and-center.

Apply now: https://apply.workable.com/j/4EDB5A3F05