Job Description

Master Works is seeking an AI Engineer with strong generative AI prototyping experience to rapidly turn ideas into working AI-powered demos, proofs of concept, and early-stage products. The role sits between research and production engineering — combining hands-on use of large language models, multimodal models, retrieval-augmented generation, and agent frameworks to validate feasibility, explore new capabilities, and de-risk feature delivery. The role is central to accelerating AI innovation across the data, analytics, and intelligence portfolio, with a particular emphasis on bilingual (Arabic/English) AI use cases.

• Rapidly design and build generative-AI prototypes and proofs of concept — from initial idea to working demo — using modern LLM APIs, open-source models, and prototyping frameworks.
• Design and implement retrieval-augmented generation (RAG) pipelines, including document ingestion, chunking strategies, embedding generation, vector storage, and retrieval orchestration.
• Build agentic systems that use tool/function calling, planning, and multi-step reasoning to automate complex tasks; integrate AI agents with internal APIs, databases, and external services.
• Engineer prompts, system instructions, and structured-output schemas; iterate on prompt design through measurable evaluation rather than guesswork.
• Work with multimodal AI capabilities — text, vision, speech (STT/TTS), and document understanding — and combine them into coherent end-to-end AI pipelines.
• Build and run AI evaluation frameworks (offline benchmarks, LLM-as-judge, golden-set regression testing) to measure quality, accuracy, latency, and cost across model and prompt variants.
• Adapt and fine-tune open-source models (LoRA, QLoRA, full fine-tuning) when proprietary or domain-specific behaviour is required, especially for Arabic-language tasks.
• Apply responsible-AI practices — safety filtering, hallucination mitigation, prompt-injection defence, PII handling, and bias awareness — especially for sensitive or regulated environments.
• Hand off mature prototypes to production engineering teams, providing clear documentation, evaluation results, deployment notes, and migration guidance.
• Demo prototypes and AI capabilities to internal stakeholders and clients; explain technical concepts, limitations, and risks to non-technical audiences.
• Track and evaluate fast-moving developments in the generative-AI ecosystem (new models, tools, techniques, and providers) and recommend adoption where they materially improve outcomes.
• Collaborate cross-functionally with product, design, data, and software engineering teams to shape AI features end-to-end and ensure they deliver real user value.

Requirements

• Bachelor’s degree in Computer Science, Artificial Intelligence, Machine Learning, Data Science, Software Engineering, or a related discipline.
• Postgraduate qualification (MSc/PhD) in Machine Learning, Artificial Intelligence, Natural Language Processing, or a related field is preferred.
• Recognised certifications in machine learning, deep learning, generative AI, or cloud AI platforms (AWS, Azure, GCP) are a strong plus.
• 3–5+ years of AI/ML engineering experience, with at least 1–2 years hands-on building generative-AI applications using modern LLMs.
• Demonstrable portfolio of shipped prototypes or production AI features — ideally including RAG systems, AI agents, or multimodal pipelines.
• Hands-on experience with leading LLM APIs (e.g., Anthropic Claude, OpenAI GPT, Google Gemini) and open-source model serving (e.g., Hugging Face, vLLM, Ollama).
• Strong Python proficiency, including modern AI libraries such as PyTorch, Hugging Face Transformers, LangChain or LlamaIndex, and standard data tooling.
• Working experience with embedding models, vector databases (e.g., Pinecone, Weaviate, Qdrant, FAISS), and retrieval evaluation techniques.
• Familiarity with rapid-prototyping tools (Streamlit, Gradio, Jupyter), API integration, and lightweight web back-ends.
• Experience designing AI evaluation harnesses (offline benchmarks, LLM-as-judge, golden sets) and using observability/tracing tools for AI systems.

• Exposure to fine-tuning techniques (LoRA, QLoRA, instruction tuning) and to Arabic-language NLP is strongly preferred.
• Working knowledge of containerisation (Docker), version control (Git), and cloud AI platforms (AWS Bedrock, Azure OpenAI, GCP Vertex AI) is a strong plus.
• Excellent leadership and team development skills.
• Strong bias for action and a rapid-prototyping mindset; able to ship a working demo in days, not months.
• Comfortable working under uncertainty — rapidly forming hypotheses, running experiments, and iterating based on evidence.
• Strong analytical and problem-solving skills, with the discipline to evaluate AI behaviour quantitatively rather than relying on intuition alone.
• Excellent verbal and written communication, with the ability to demo AI capabilities and explain technical concepts, limitations, and risks to non-technical audiences.
• Curiosity and continuous-learning mindset; actively follows the rapidly evolving generative-AI landscape and brings new ideas back to the team.
• Collaborative team player; comfortable in cross-functional teams of researchers, software engineers, designers, and client business owners.
• Cultural awareness and sensitivity, particularly when designing AI experiences for Arabic-speaking users in regional contexts.