AI Architect - AWS Hybrid AI Platform, Security and LLMOps
NTT DATA Romania SA · Posted Nov 12, 2026
- Cluj, Romania
- Hybrid
- Full-time · Lead
- 10–10 years
Design and govern secure, production-grade AI platform and solution patterns across an AWS-centered hybrid ecosystem, focusing on platform integration, runtime patterns, private connectivity, identity, model gateways, managed and open-source models, LLMOps/MLOps, observability, reliability, cost, security and responsible AI.
Responsibilities
- Own reusable platform and solution architecture patterns for AI applications, agents, models, retrieval services, data flows, APIs and hybrid infrastructure.
- Architect containerized, serverless, event-driven and private model-serving workloads using Kubernetes, EKS/EKS Anywhere, ECS, Lambda and approved runtime platforms.
- Design secure cloud-to-on-premises connectivity using Direct Connect, Storage Gateway, enterprise networking, private endpoints, DNS, proxies and reverse-proxy patterns.
- Design identity and access controls using AWS IAM, Secrets Manager, Active Directory integration, workload identities, least privilege, encryption, certificates and audit logging.
- Define model-gateway, model-routing, open-source model, managed-model and private inference patterns, including safety, tenancy, quotas and cost controls.
Requirements
- Nice to have: regulated environments