Job description
SentraAI is hiring a mid-level GenAI Engineer to join a dedicated delivery team embedded with a major energy sector client in Abu Dhabi.
The role focuses on building production-grade generative AI capabilities, including Microsoft Copilot extensibility, Azure OpenAI powered assistants, Retrieval-Augmented Generation (RAG) pipelines, and Retrieval-Augmented Fine-Tuning (RAFT) workflows, that put domain-aware AI in the hands of the client's engineering, operations, and corporate users.
You will work hands-on across the full GenAI delivery lifecycle, covering data ingestion and indexing, prompt design, evaluation, deployment, and ongoing optimisation, in close partnership with senior SentraAI architects and client stakeholders.
About SentraAI SentraAI is a specialist enterprise AI firm focused on production-grade delivery in regulated environments.
We design, build, and operationalise AI systems for clients across banking, finance, energy, and government sectors in the GCC and beyond.
Our work is defined by engineering rigour, deep domain partnership, and a commitment to solutions that perform reliably at enterprise scale.
• Production-grade AI, not prototypes.
Our engagements ship into regulated, mission-critical environments, so what you build will be used by real operators every day.
• Embedded with industry-defining clients.
This role places you at the heart of the GCC energy sector, working on problems that move the dial for a global business.
• Deep technical mentorship.
You will work alongside senior SentraAI architects and engineers with extensive enterprise AI delivery experience.
• Modern GenAI stack.
Hands-on exposure to Azure OpenAI, Microsoft Copilot, advanced retrieval, RAFT, and evaluation tooling, kept current with the rapid pace of the field.
• Growth pathway.
Clear progression into Senior Engineer, GenAI Architect, or Delivery Lead tracks as you build delivery breadth.
• A team that values craft.
Engineering rigour, clear thinking, and quality of output matter here, and we hire people who care about doing the work properly.
GenAI Solution Engineering • Design and implement RAG pipelines that index structured and unstructured client content, including technical manuals, operational records, SOPs, knowledge bases, and ticketing data, and surface it through chat, search, and Copilot experiences.
• Develop and tune Retrieval-Augmented Fine-Tuning (RAFT) workflows where standard RAG is insufficient, balancing retrieval quality with model adaptation for domain-specific tasks.
• Build Microsoft Copilot extensions, Copilot Studio agents, and Azure OpenAI backed assistants that integrate with Microsoft 365, Teams, SharePoint, and client line-of-business systems.
• Implement prompt engineering patterns, structured output handling, function and tool calling, and grounding strategies that produce reliable, auditable outputs for technical and operational users.
Data, Retrieval and Evaluation • Build and maintain document ingestion pipelines, chunking and embedding strategies, vector indexing, and hybrid retrieval (semantic plus keyword) using Azure AI Search or equivalent.
• Establish evaluation harnesses covering accuracy, groundedness, latency, cost, and safety, with golden datasets and automated regression testing on every release.
• Diagnose retrieval and generation failures through systematic error analysis, and iterate on chunking, prompts, retrieval parameters, and where appropriate, fine-tuning data.
Production Delivery • Ship code that meets SentraAI's standards for production-grade delivery, namely source-controlled, tested, observable, and deployed through Azure DevOps pipelines.
• Implement guardrails, content filtering, PII handling, and audit logging consistent with the client's information security and compliance requirements.
• Monitor cost, latency, and quality of deployed GenAI services and tune accordingly.
Collaboration and Onsite Delivery • Work onsite at the client alongside business stakeholders, IT and architecture teams, and SentraAI delivery leads.
• Partner with SentraAI Azure Cloud, DevOps, and Programme Management colleagues to deliver end-to-end solutions.
• Document architecture, design decisions, and operational runbooks for handover and ongoing support.
Required Qualifications Experience • 3 to 5 years of professional software engineering experience, including at least 1 to 2 years building production GenAI, LLM, or applied ML solutions.
• Demonstrable delivery of at least one production-grade RAG application, end-to-end, from data ingestion to user-facing experience.
• Hands-on experience with Azure OpenAI Service or comparable LLM platforms in an enterprise setting.
Technical Skills • Strong Python skills, with familiarity in C# or TypeScript a plus given Copilot and Microsoft 365 extensibility work.
• Working knowledge of Azure AI Search, or an equivalent vector store such as pgvector, Pinecone, Weaviate, or OpenSearch, plus embedding models and hybrid retrieval techniques.
• Practical experience with prompt engineering, function and tool calling, structured outputs, and LLM evaluation methods.
• Familiarity with Microsoft Copilot extensibility, covering Copilot Studio, declarative or custom agents, and M365 connectors, or demonstrated ability to ramp quickly.
• Exposure to RAFT or other retrieval-augmented fine-tuning approaches, with comfort in SFT or LoRA style adaptation welcomed.
• Comfort working with Azure DevOps, Git, CI/CD, containers, and basic Infrastructure-as-Code such as Bicep or Terraform.
Ways of Working • Track record of shipping into regulated or enterprise environments where reliability, security, and auditability are non-negotiable.
• Strong written and spoken English, with the ability to communicate technical concepts to non-technical stakeholders.
• Comfortable working onsite at a client location and collaborating across mixed SentraAI and client teams.
Advantageous but Not Mandatory • Prior exposure to oil and gas, energy, or industrial domains and the associated document and data types.
• Experience with LangChain, LlamaIndex, Semantic Kernel, or comparable orchestration frameworks.
• Familiarity with multi-agent or agentic patterns, including planning, tool use, and the Model Context Protocol.
• Microsoft certifications such as AI-102 (Azure AI Engineer Associate), AZ-204, or AI-900.
• Background contributing to evaluation frameworks such as Ragas, Promptfoo, or Azure AI Studio evaluations.
• Arabic language skills.
This job post has been translated by AI and may contain minor differences or errors.
