Senior AI & Data Platform Owner

We are seeking a Senior IT platform professional to act as the AI & Data Platform Owner within DIGITALcore responsible for the vision, roadmap, and hands-on delivery of GDMS’ secure AI, GenAI, and data platform capabilities in AWS. You will build “paved roads” that enable GDMS programs and internal teams to deploy AI/ML and data workloads quickly, securely, and repeatably,aligned to regulated-environment requirements.

This role blends deep platform and infrastructure expertise with service ownership thinking. You will own the platform backlog, define reference architectures, lead implementation, and ensure operational excellence. The ideal candidate is customer-obsessed and UX-minded, able to translate real user workflows into secure, intuitive IT platform experiences.

Key Responsibilities

AI & Data Platform Ownership

• Own the AI & Data platform roadmap inside DIGITALcore: define priorities, epics, and release plans for GenAI, ML, and data foundations in AWS.
• Establish platform outcomes and metrics: onboarding speed, model deployment cycle time, reliability, security posture, cost efficiency, and user satisfaction.
• Align priorities with DIGITALcore governance, enterprise architecture, security/compliance, and program delivery needs.

GenAI Platform (Amazon Bedrock)

• Lead the secure deployment and operationalization of Amazon Bedrock capabilities (model access, guardrails, logging, governance, and cost controls).
• Build reusable patterns for GenAI applications: prompt management, RAG patterns, embeddings, evaluation, and production-grade integration patterns.
• Drive implementation and adoption of Bedrock AgentCore and Strands for agentic workflows, orchestration patterns, tools/functions, and safe execution in a controlled environment.
• Define “secure GenAI paved roads” including identity/access patterns, data access boundaries, network controls, and auditability.

ML Platform & MLOps (Amazon SageMaker)

• Lead the deployment of GDMS’ MLOps capability using Amazon SageMaker and related AWS services.
• Establish standard pipelines for training, evaluation, model registry, approval gates, packaging, deployment, monitoring, and rollback.
• Implement and maintain reference architectures for:
  • Feature engineering and feature store patterns (as applicable)
  • Model training workflows (batch/stream)
  • Real-time and batch inference architectures
  • Model monitoring (drift, bias, performance) and retraining triggers
• Enable program teams with templates, starter kits, and repeatable CI/CD patterns for ML workloads.

Data Platform Foundations

• Define and evolve secure data platform patterns under DIGITALcore: data ingestion, storage, cataloging, governance, and access controls.
• Build secure, compliant data access patterns to support analytics and AI workloads (including data classification handling and least-privilege access).
• Ensure alignment with enterprise logging, monitoring, and audit evidence requirements for data movement and model usage.

Secure Cloud Construct & Compliance Enablement

• Operate within DIGITALcore’s secure landing zone constraints: segmentation, encryption, identity controls, logging, and evidence generation.
• Translate regulated-environment requirements into platform guardrails and automation (secure-by-default architectures).
• Build audit-ready artifacts: standard configs, platform runbooks, automated checks, control mappings, and evidence packages.

Operational Excellence & Customer Experience

• Own platform reliability: SLOs, incident management, change control, and operational runbooks.
• Treat internal users as customers: conduct user discovery, map workflows, reduce friction, and improve “time-to-first-model” and “time-to-first-agent.”
• Improve usability with clear docs, onboarding paths, templates, and opinionated “golden paths” that teams can follow safely.

Cross-Functional Leadership

• Partner with security, architecture, and program engineering teams to ensure platform capabilities meet mission needs.
• Influence standards for AI governance (usage monitoring, guardrails, model approval gates, and lifecycle controls).