Job details:

• Role: Lead AI Platform Architect
• Full-time
• Remote
• Start date: 15/06/2026
• End date: 14/06/2027

Scope:

• Run Weekly Architecture Review Boards
  Evaluate HLDs/LLDs from deployment teams, validate solution patterns (Databricks, AI Foundry, lakehouse, model registry, feature store), ensure alignment with enterprise standards, and provide architectural approvals or required redesigns.
• Provide Deep Dive Guidance & Technical Workshops
  Host weekly topic-focused sessions (e.g., private networking, identity patterns, MLOps pipelines, model serving, GPU/CPU infra, agentic AI workflows) to upskill deployment teams and guide them through complex architectural decisions.
• Review and Govern Identity, Security & Compliance
  Audit upcoming deployments for correct IAM/RBAC/Managed Identity/OIDC usage, ensure zero trust principles, validate private endpoint designs, and identify risks like privilege creep, misconfigured networking, or unsafe model/data handling.
• Support Deployment Teams on Complex Blockers
  Provide hands on architectural consultancy for issues involving data plane access, cluster policies, lineage/quality gaps, CI/CD failures, inference latency, scaling problems, or governance conflicts. Offer design alternatives, not tactical fixes.
• Maintain & Evolve Architecture Templates and Reference Patterns
  Update reusable reference architectures, IaC patterns, naming conventions, cluster policies, identity patterns, and AI deployment standards. Ensure all documentation reflects current best practices and is consumable by delivery teams.
• Perform Weekly Governance & Cost Health Oversight
  Review platform wide usage: cluster cost anomalies, storage growth, model serving load, resource distribution, and architectural drift. Provide a weekly report to leadership and deployment teams with required remediations and improvement actions.

Skills:

• Databricks (Azure/AWS): Leads regulated builds (HIPAA/PCI), multi account, delta sharing, strong guardrails – must have knowledge of Databricks
• AI Infra (Training & Serving): Optimizes packings; multi model endpoints, caching, eval loops – must have
• Azure AI Foundry / AWS AI: Standardizes model ops, guardrails, tenancy patterns – nice to have
• MLOps & CI/CD: Org level reusable workflows, provenance/SBOM

SNI sp. z o.o. will process personal data for the purpose of the recruitment process in accordance with Data Privacy Policy. The data may also be stored and processed for future recruitment purposes, in accordance with the given consent.