Multi-Cloud AI Architect

About the role

Design AI platforms that span AWS, Azure, and GCP without becoming lowest-common-denominator mush. On Nebulai contract engagements, you help clients pick where workloads should live: Bedrock vs Azure OpenAI vs Vertex, shared eval/observability patterns, identity and networking across clouds, and escape hatches when a vendor or region constraint appears. Marketplace talent advising multi-cloud AI delivery—not claiming employment at any hyperscaler.

Responsibilities

• Design AI/ML solutions that operate across AWS, Azure, and GCP.
• Guide clients in selecting appropriate workload placements (e.g., Bedrock, Azure OpenAI, Vertex AI).
• Establish shared evaluation and observability patterns across clouds.
• Implement identity, networking, and key‑management strategies that span multiple providers.
• Create escape‑hatch plans for vendor or region constraints.
• Act as marketplace talent advising on multi‑cloud AI delivery.

Requirements

• Designed or delivered AI/ML solutions on at least two major cloud platforms.
• Able to explain trade‑offs (model access, data residency, cost, operational burden) without dogma.
• Knowledge of enterprise constraints: SSO, private networking, key management, logging.
• Capable of leaving behind reference architectures that teams can actually implement.

Skills (Bonus)

• Experience building abstraction layers or gateway patterns for multi‑provider LLMs.
• Center‑of‑Excellence or platform experience standardizing practices across clouds.
• Regulated or global deployment experience.