QA Automation Engineer

MICLA Engineering & Design was established in 2005 with the aim of supporting its clients by providing qualified engineering and software development resources in institutional and enterprise environments. The company operates in mission‑critical projects with high requirements in terms of quality, reliability, security and long‑term maintainability. Position: Workplace: Tunis (Tunisia)
- On‑site / Hybrid Position required: 5 open positions Job title: SW GenAI / LLM Engineer — Automotive Software Engineering

About The Role
A major international vehicle manufacturer is embedding generative AI into the way its software organisation works — not as a side experiment, but as a required capability across development, testing and quality engineering. Micla Engineering & Design is building the team that delivers it. This is not a research position and not a chatbot position. You will design, deploy and optimise LLM-based and agentic systems that operate on real engineering artefacts: requirements, test cases, code, review findings, release documentation. The output has to be correct, traceable and auditable, because it feeds a safety-critical development process. If you want to work on agentic systems where the evaluation problem is as hard as the generation problem, this is that role. What You Will Do Design, deploy and optimise generative AI and large language model systems for software engineering workflows Architect agentic systems: orchestration, tool integration, task decomposition Build retrieval and grounding strategies (RAG) over engineering artefacts — requirements, specifications, test documentation, code repositories Design prompts and structured output schemas that produce machine-verifiable results Build the evaluation layer: correctness benchmarks, regression suites for model behaviour, safety and guardrail testing Operate the LLMOps pipeline — versioning, monitoring, cost and latency management, drift detection Integrate AI-assisted steps into existing CI/CD pipelines with proper quality gating

Requirements:
Required Engineering degree in Computer Science, Software Engineering, Data Science, Electrical Engineering or related field Strong Python Practical experience building applications on top of LLM APIs or open-weight models — you have shipped something, not only read about it Understanding of retrieval-augmented generation: embeddings, vector stores, chunking strategies, grounding and citation Prompt and output-schema design, with structured formats (JSON schema, function calling, tool use) Git, and experience integrating work into CI/CD pipelines A rigorous approach to evaluation — you can explain how you know a system got better, with numbers Professional English, written and spoken Valued Agent frameworks and orchestration tooling; multi-step tool-using agents LLMOps: observability, tracing, evaluation harnesses, prompt versioning Fine-tuning, adapters, or model distillation CI/CD platforms: Jenkins, GitLab CI, TeamCity; multi-stage pipelines with gating and approvals Software quality tooling and standards: static code analysis, MISRA, unit testing, coverage Requirements management tooling (DOORS Next Generation or equivalent) and ALM systems Awareness of safety-critical or regulated development processes, and of what auditability demands from an AI system Agile delivery experience Seniority levels open Junior (0-3 years) Mid (4-6) Advanced (7-9) Expert (10+). This is a young field and we know it. A Junior candidate with a serious portfolio of built projects will be considered ahead of a senior profile with none. Send us what you have built.