Enterprise Architect – Data & AI

Profile: The Enterprise Architect – Data & AI provides strategic technology leadership across data platforms and AI systems - defining target architectures, accelerating delivery through reusable patterns, and building stakeholder confidence by aligning with modern regulations and standards. The focus spans generative AI, agentic AI, retrieval augmented architectures, and governed, secure deployment at enterprise scale across Microsoft Azure, AWS, and Google Cloud.

Working closely with federated Data & AI teams, the role establishes an AI‑ready foundation, ensuring compliant, secure, and reliable integration with enterprise platforms such as SAP, Workday, Salesforce, Jaggaer, and ServiceNow, and with multi‑cloud data/AI services (Azure, AWS, Google Cloud).

Key Responsibilities

• Enterprise Data & AI Strategy and Roadmap: Through partnership, drive the multi‑year strategy and modernisation roadmap; define target state and incremental evolution aligned to business value and regulatory timelines (e.g., EU AI Act milestones).
• Architecture Patterns (MultiCloud): Establish and govern patterns for AI on enterprise data: RAG, hybrid search (vector + keyword), prompt orchestration, grounding/attribution, agentic AI architectures and standards such as Model Context Protocol (MCP) - implemented across:
  • Microsoft: Azure OpenAI Service, Azure AI Search, Azure Cosmos DB (MongoDB vCore) vectors, Microsoft Fabric Real‑Time Intelligence.
  • AWS: Amazon Bedrock (model orchestration & safety), Amazon SageMaker (MLOps), Amazon Kendra/Athena/OpenSearch (search/query), Amazon DynamoDB/Neptune/DocumentDB (storage), AWS Glue/Lake Formation (data catalog/permissions), Amazon MSK/Kinesis (streaming).
  • Google Cloud: Vertex AI (model lifecycle, evaluation), Generative AI Studio, Vector Search in BigQuery/AlloyDB, Cloud SQL/Spanner/Firestore/Neo4j Aura (graph), Dataflow/Pub/Sub (streaming), Dataplex/Data Catalog (governance).
• LLMOps/MLOps & DevOps: Define practices for model and prompt versioning, systematic evaluation, content safety, monitoring (latency, quality, cost), rollback, and incident response; embed Responsible AI controls across Azure, AWS, and GCP equivalents (e.g., Bedrock Guardrails, Vertex AI evaluation).
• AI Governance & Compliance: Implement AI governance frameworks and controls mapped to NIST AI RMF and ISO/IEC 42001, and prepare for EU AI Act obligations (including GPAI considerations), with consistent policy enforcement across clouds and enterprise platforms.
• Secure ‘AI on Your Data’ (Zero‑Trust by Design): Architect solutions that keep data protected end‑to‑end: encryption, confidential/secure computing options, identity & access controls, DLP, data residency, and policy enforcement. Leverage:
  • Azure: Azure OpenAI Service, Azure AI Search, Cosmos DB vectors, Fabric OneLake.
  • AWS: Bedrock, SageMaker, Kendra/OpenSearch, Lake Formation with fine‑grained permissions.
  • GCP: Vertex AI, BigQuery with vector search, Dataplex governance.
• Real‑Time & Streaming Analytics: Lead real‑time architectures to enable AI‑driven decisions and automation:
  • Azure Event Hubs, Stream Analytics, Fabric Real‑Time Intelligence.
  • AWS Kinesis (Data Streams/Firehose), Amazon MSK (Apache Kafka).
  • GCP Pub/Sub, Dataflow (stream/batch).
• Data Organisation, Serving & Federated Governance: Define lakehouse (Delta/Iceberg), data mesh, and federated governance patterns; balance decentralisation with central standards and shared services across the three clouds. Establish architectural standards for secure and efficient data serving to end users and applications, ensuring compliance and performance.
• FinOps for AI: Partner with Finance to set cost guardrails, forecasting, and optimisation across inference, storage, streaming—standardising FinOps practices and dashboards for Azure, AWS, and GCP workloads.
• Research & Foresight: Conduct extensive, continuous research on emerging AI models, vector databases, observability, prompt safety, privacy‑enhancing technologies, and regulatory developments across Microsoft, AWS, and Google Cloud. Produce future‑ready design recommendations, evaluate design trade‑offs, run comparative pilots/PoCs, and publish reference blueprints that anticipate next‑gen capabilities and standards.
• Enablement & Communities of Practice: Coach and upskill engineering and delivery teams; establish communities of practice; create reference blueprints, playbooks, and training aligned to enterprise standards.

Skills / Requirements

• Bachelor’s degree in Computer Science or related field; advanced degree beneficial.
• 10+ years in data architecture/platform engineering (including 5+ years designing end‑to‑end data/analytics solutions; 2+ years leading GenAI/LLMOps programmes).
• Deep expertise with Microsoft Azure and Microsoft Fabric (OneLake; Data Engineering/Factory; Warehouse; Real‑Time Intelligence) and integration into M365.
• Hands‑on experience with AI on enterprise data (RAG, embeddings, vector search) using:
  • Azure: Azure OpenAI Service, Azure AI Search, Cosmos DB vectors.
  • AWS: Amazon Bedrock, Amazon SageMaker, Amazon Kendra/OpenSearch, DynamoDB/Neptune/DocumentDB.
  • Google Cloud: Vertex AI, BigQuery/AlloyDB vector search, Generative AI Studio.
• Strong understanding of data storage paradigms (Delta/Iceberg lakehouse, relational, document, graph) and streaming architectures (Event Hubs/Kinesis/Pub/Sub).
• Proven knowledge of data governance tools for cataloguing, classification, lineage, data quality, and policy enforcement across Azure, AWS, and GCP.
• Security & privacy expertise: content safety, prompt/indirect injection mitigations, encryption, identity/access, and confidential computing patterns across cloud AI services.
• LLMOps/MLOps & DevOps: CI/CD for models and prompts, evaluation frameworks, observability (including OpenTelemetry where applicable), rollback and incident runbooks.
• Familiarity with EU AI Act, NIST AI RMF, ISO/IEC 42001; ability to translate requirements into technical and process controls across clouds.
• Excellent communication; ability to translate between business outcomes and technical execution; proven stakeholder influence at senior/executive levels.

Preferred Certifications

• TOGAF or equivalent enterprise architecture certification.
• Microsoft Certified: Azure Solutions Architect Expert; Azure Data Engineer Associate; Fabric Analytics Engineer Associate.
• Other relevant certifications is an added advantage:
  • AWS Certified: Solutions Architect – Professional; Machine Learning – Specialty; Data Engineer – Associate.
  • Google Cloud Certified: Professional Cloud Architect; Professional Machine Learning Engineer; Professional Data Engineer.
  • Security/privacy certifications (e.g., CIPP/E) beneficial., playbooks, and training aligned to enterprise standards.