AI Engineer

The Institute

AI and Digital Science Research Centre

Duties & Responsibilities

• Design and build agentic AI systems (multi-agent workflows, tool use, planner and executor patterns) on top of Falcon and other foundation models, applied to enterprise and government use cases.
• Architect end-to-end RAG pipelines, covering ingestion, chunking, embedding, hybrid retrieval, reranking, and grounded generation across structured and unstructured data sources.
• Own model serving and inference optimization with vLLM (or equivalent frameworks), including batching, KV caching, quantization, and throughput tuning to meet latency and cost targets.
• Lead prompt engineering, model selection, and lightweight fine-tuning (PEFT, LoRA, SFT) where it measurably improves application quality.
• Build rigorous evaluation frameworks and online and offline benchmarks to measure quality, safety, latency, and cost of deployed AI systems, and act on the results.
• Investigate and mitigate AI failure modes such as hallucination, retrieval gaps, prompt injection, and unsafe outputs, and codify the fixes into reusable patterns.
• Partner with full-stack engineers, researchers, and product teams to translate business problems into well-scoped AI solutions, then ship them to production and iterate based on real usage.
• Contribute to internal AI frameworks, libraries, and engineering best practices that accelerate future deployments across the centre.

Job Specification

• Bachelor’s or Master’s degree in Computer Science, Artificial Intelligence, Machine Learning, Software Engineering, or a related discipline.
• 2 to 8 years of hands-on experience building and deploying AI or machine learning systems in production environments.
• Strong Python skills, with deep experience in modern AI and ML libraries such as PyTorch, Hugging Face Transformers, LangChain or LlamaIndex, and serving frameworks such as vLLM or TGI.
• Proven experience with agentic frameworks and design patterns (function calling, tool use, multi-agent orchestration) and with building production RAG systems.
• Hands-on experience with model serving and inference optimization, including batching, KV caching, and INT4/INT8 quantization for LLMs.
• Practical experience with prompt engineering, model evaluation, and lightweight fine-tuning (PEFT, LoRA, or SFT) of foundation models.
• Solid software engineering foundations, including Python API development, version control, testing, and code review.
• Working knowledge of containerization (Docker), CI/CD, and basic MLOps practices such as model and prompt versioning and observability.
• Familiarity with vector databases (e.g., Milvus, Qdrant, pgvector, Weaviate) and embedding-based retrieval at scale.
• Strong communication skills and the ability to engage directly with internal stakeholders and enterprise clients in English (Arabic is a plus).

Additional Skills

• Patents.
• Publications in top-ranking conferences and journals, such as ICML, NeurIPS, ICLR, KDD, VLDB, AAAI, and ACL, to cite a few.
• Experience deploying AI systems in regulated industries (government, financial services, defence) or in Arabic-language contexts.
• Contributions to open-source AI projects.