Role type - Permanent / Fixed Term / Contracting :
Contracting
Mode of working Hybrid / office based :
Hybrid
If Hybrid, how many days are required in office? :
3 days
The Role
As an Azure AI Architect, you will lead the end to end architecture, design, and delivery of enterprise AI solutions built on Azure AI Foundry, Azure OpenAI, Azure AI Document Intelligence design and deploy. You should have hands-on experience in Azure Data Factory, Azure ML Service, Azure Data bricks, Azure Key Vault, Azure BLOB Storage Account, Azure Log Analytics, Synapse Analytics, Azure DevOps. You will define the reference architectures and guardrails that enable teams to build secure, scalable, and cost efficient AI capabilities-spanning intelligent document processing, generative AI applications, and retrieval augmented systems integrated with core enterprise platforms.
You will collaborate closely with business stakeholders, product owners, security and compliance, data engineering, and application teams to translate business outcomes into actionable AI roadmaps and solution designs. You will also guide implementation teams on best practices for prompt engineering, grounding strategies, RAG patterns, observability, MLOps, and Responsible AI-ensuring solutions are robust, compliant, and production ready.
This role suits a hands on architect who can balance strategic vision with pragmatic delivery, establish engineering standards, and mentor developers while owning architectural decisions and non functional requirements (performance, resilience, security, and cost).
Your Responsibilities
Define target state Azure AI reference architectures leveraging Azure AI Foundry, Azure OpenAI, and Azure AI Document Intelligence for enterprise use cases (IDP, GenAI copilots, knowledge retrieval).
Establish standards for RAG architectures (vectorization, indexing, chunking, grounding, citation policies) using Azure Search or vector enabled data stores.
Create architectural decision records (ADRs), guardrails, and reusable blueprints for solution teams.
Lead end to end solution designs including API contracts, integration patterns (Azure Functions, Logic Apps, Event driven), security boundaries, and observability.
Architect document intelligence pipelines (classification / extraction / OCR / validation) and integrate with downstream systems (CRM / ERP / ITSM / EDM).
Define non functional requirements (availability, latency, throughput, cost, DR / RTO RPO) and ensure solutions meet them.
Implement identity & access (Entra ID), data isolation, Key Vault secrets, network security (Private Endpoints), and content filtering.
Embed Responsible AI practices : safety filters, prompt / content governance, data privacy, red teaming guidance, and human in the loop review where needed.
Ensure regulatory alignment (e.g., GDPR, ISO controls) and collaborate with risk, legal, and security for approvals.
Define MLOps processes for versioning, evaluation, promotion, rollback, and monitoring (latency, cost, drift, hallucination rate, safety events).
Instrument observability (logging, tracing, metrics) and error budgets; establish SLIs / SLOs for AI services.
Drive FinOps discipline-capacity planning, token / cost controls, caching strategies, and model selection / optimization.
Mentor engineers on prompt engineering, grounding, tool use patterns, and quality evaluation frameworks.
Your Profile Essential skills / knowledge / experience
15+ years overall in software / solution architecture with 7+ years in cloud (Azure) and 5+ years in AI / ML or Azure AI solutioning.
You should have hands-on experience in
Deep knowledge of LLMs (prompting, system prompts, grounding, evaluation), embeddings, and RAG design (index selection, chunking strategies, reranking, citations).
Strong design of secure API and event driven integration patterns; familiarity with microservices and domain driven design concepts.
Practical expertise in Entra ID, RBAC, network isolation (Private Links), secret management, data residency, and Responsible AI controls.
Experience aligning solutions with privacy and compliance requirements (e.g., GDPR, ISO 27001 controls) and completing architecture risk assessments.
Proficiency in Python and / or C# for service integration, evaluation tooling, and adapters; solid understanding of REST APIs, JSON, and async patterns.
Experience with vector databases (Cosmos DB with vector search, Redis Enterprise, or Pinecone) and semantic ranking.
Knowledge of Azure Machine Learning for training / evaluation pipelines and model registries.
Familiarity with Power Platform (AI Builder, Power Automate) for rapid AI-enabled workflows.
Exposure to multi agent / tool use orchestration, guardrails frameworks, and evaluation harnesses for GenAI.
Performance engineering (token optimization, caching, partial responses / streaming) and cost to serve modeling.
Certifications
#J-18808-Ljbffr