Full Stack AI Engineer

Full Stack AI Engineer — Jordan (Remote) Company: AllysAI (Tuitify FZ LLC) Location: Fully remote, based in Jordan Type: Full-time Level: Mid-level (3–5 years) Reports to: Head of AI Compensation: Competitive fixed salary, benchmarked to experience About AllysAI We rebuild enterprise operations as AI-native systems. Production AI shipped like products, powered by AllysOS. Our clients are large enterprises — pharma, banking, logistics, energy, government — across the Gulf, Europe and Latin America. What we build goes into production and gets used by real operators every day. We are not a demo shop and we do not ship prototypes and call them products. Headquartered in Abu Dhabi, with offices in New York and Lausanne, and engineering that works wherever the talent is. The role "Full stack" here means full stack across the AI system — not React and a database. You will own a workload end to end: getting the data into shape, building the retrieval and agent logic on top of it, building the evaluation harness that proves it actually works, and shipping it into production where it stays reliable under real client load. Data in at one end, a system a client trusts at the other. If you have spent the last two years wiring an LLM into a chat box and calling it done, this is not that. We care most about the part that most people skip: knowing whether the thing works, and being able to prove it. What you'll do Data. Build the pipelines that get client data — documents, records, exports from systems that were never designed to be read by machines — into a state a model can actually use. Chunking, extraction, cleaning, structuring, indexing.LLM and agent engineering. Design and build RAG systems, multi-step agents, tool use, and orchestration. Choose the right approach for the problem rather than reaching for the same pattern every time.Evaluation. Build eval sets and harnesses for the systems you ship. Define what "correct" means for a given client workload, measure it, catch regressions before the client does. This is a first-class part of the job, not an afterthought.Production. Deploy, monitor, and keep it running — latency, cost per call, failure modes, observability. Own what you ship after it ships.Iterate on real feedback. Sit close to how clients actually use the system and improve it based on what happens in production, not what happened in your test set. What we're looking for3–5 years building software, with meaningful recent time spent on LLM-based systemsStrong Python — this is the working language of the roleReal experience with LLM application engineering: RAG, embeddings and vector search, agent frameworks, tool calling, structured output, prompt engineering as an engineering discipline rather than a guessing gameEvaluation experience — you have built eval sets, run systematic testing on model outputs, and can talk about how you measured quality rather than eyeballed itData engineering ability — you can handle messy real-world data end to end without waiting for someone to hand you a clean datasetProduction experience — Docker, CI/CD, cloud (AWS, Azure or GCP), API design. You have shipped things that other people depend on.Excellent English, written and spoken — you will read client documentation and write for an international teamBased in Jordan and able to work with meaningful overlap with Gulf hoursIndependent. Remote means you unblock yourself, communicate clearly in writing, and do not need to be chased. Nice to haveFine-tuning, distillation, or model optimisation experienceEnterprise integration work — ERPs, CRMs, internal APIs, legacy systemsFront-end ability (React/Next.js) for building interfaces over what you constructArabicOpen-source contributions or work we can look at What we offerFully remote, working from Jordan, on a permanent full-time basisCompetitive fixed salary benchmarked to your experienceReal enterprise problems at scale — not internal tooling, not another chatbotA small, senior engineering team where your work is visible and your decisions matterDirect exposure to how AI is actually being deployed inside large organisations in the Gulf and EuropeRoom to grow into technical ownership of a product line as we scale