Azure AI Engineer / Fixed-Term Contract / Remote, UK Based / Competitive compensation
About the Company:
The company is an AI-focused business building practical AI-native solutions for clients, with a strong focus on automation, intelligent workflows and applied AI products that solve real business problems.
The business is now looking for a hands-on Azure AI Engineer / AI Agent Developer to support the delivery of an Azure-based AI agent and RAG workflow, while also helping build internal AI-native automations across commercial, product and operational workflows.
About the Role:
This is a fixed-term contract role (August & September 2026) suited to a practical Azure AI builder who can configure services, write code, debug issues and improve the quality of AI agent and retrieval-based systems.
The role will be approximately 70% focused on client delivery and 30% focused on internal automation and light web support. The main project will involve building, testing, monitoring and improving an Azure-based AI agent / RAG workflow using Azure AI Search, Cosmos DB, Azure AI Foundry, Azure OpenAI, Application Insights and Azure Monitor.
Alongside this, the contractor will support with Power Automate workflows, AI-native automations and simple TypeScript website edits when required.
Key Responsibilities:
- Build and improve an AI underwriting assistant / agent using Azure AI Foundry, Azure OpenAI, Azure AI Search and Cosmos DB.
- Test the AI agent against client datasets and improve prompts, retrieval quality, schemas, grounding, validation and guardrails.
- Configure and maintain RAG workflows, including indexing, chunking, embeddings, vector/hybrid search and metadata filtering.
- Instrument the system using Application Insights, Azure Monitor and Log Analytics to track errors, latency, usage, retrieval diagnostics and agent performance.
- Debug agent behaviour, retrieval failures, workflow issues and common RAG failure modes.
- Work safely and effectively within client-controlled Azure environments.
- Create concise handover documentation for client and internal teams.
- Build internal AI-native automations for sales, follow-up, demos, reporting, research and product workflows.
- Build Power Automate cloud flows across Microsoft 365, Outlook, SharePoint and Teams.
- Make simple TypeScript / React / Next.js-style website edits when needed, including links, content changes, GitHub commits or pull requests.
Essential Requirements:
- Strong hands-on experience with Azure infrastructure, including resource groups, subscriptions, tenants, RBAC, managed identities, storage, Functions/serverless, secure setup and debugging in Azure environments.
- Experience building RAG workflows using Azure AI Search or equivalent vector/hybrid search systems.
- Strong understanding of indexing, chunking, embeddings, vector search, hybrid search, metadata filters, grounded answers and retrieval testing.
- Experience with Azure AI Foundry and/or Azure OpenAI, including model deployments, prompts, agent configuration, tool/function calling, structured outputs, evaluation and cost/latency awareness.
- Experience using Cosmos DB for NoSQL data modelling, including agent state, conversation history, metadata, document records, querying, indexing and partitioning basics.
- Experience with monitoring and observability tools such as Application Insights, Azure Monitor, Log Analytics and custom telemetry.
- Practical experience building AI agents, tool-using workflows, multi-step automations, orchestration, guardrails, validation and human-in-the-loop review.
- Experience creating Power Automate cloud flows across Microsoft 365, Outlook, SharePoint and Teams.
- Basic TypeScript/web experience, ideally with simple React or Next.js website edits, GitHub commits, pull requests and deployment awareness.
Ideal Candidate:
- Practical Azure AI builder, not a pure ML researcher.
- Comfortable configuring Azure services, writing code, debugging failures and improving retrieval/model quality.
- Able to work across client delivery, internal automation, product support and light web edits.
- Comfortable operating in ambiguous startup environments where requirements may evolve quickly.
- Strong communicator who can document work clearly and hand over safely to clients or internal stakeholders.
- Commercially aware and able to build AI workflows that are reliable, useful and practical for real business use cases.
Nice to Have:
- Azure Document Intelligence, OCR or PDF extraction experience.
- Experience with Durable Functions, Logic Apps or workflow orchestration.
- Dynamics 365, Dataverse or CRM integration experience.
- Experience with Semantic Kernel, LangChain or LlamaIndex.
- AI evaluation experience, including golden test sets, regression tests, retrieval quality checks and hallucination checks.
- Experience working with sensitive UK/EU client data and understanding of security/governance requirements.
…