Cognizant - Enterprise AI Platform Engineer - AWS

Level : SA and M.

Enterprise AI Platform Team Job Description :

The Enterprise AI Platform Team builds and maintains the secure, scalable AWS foundation for agentic AI workloads. This team owns infrastructure provisioning, gateways, security, and observability to enable self-service agent development.

Key Responsibilities :

- Provision and manage core AWS infrastructure including Bedrock Agent Core runtime and CSP AWS services

- Implement identity management with IAM, Cognito, Okta integration for enterprise SSO

- Build platform components : memory stores, policy enforcement engines, browser tooling

- Deploy and operate gateways : MCP Gateway (AgentCore/Kong), LLM Gateway (multimodel routing), Bedrock integration

- Manage containers and compute : Docker builds, ECR registries, EKS clusters, EC2 instances

- Author IaC with Terraform modules and maintain Jenkins/Terraform Enterprise CI/CD pipelines

- Set up RAG data infrastructure using PostgreSQL and Redshift

- Develop MCP server tools for APIs and browser integrations

- Enable multi-LLM support (AWS Nova, OpenAI, Anthropic Claude/Opus, Gemini)

- Implement comprehensive observability : CloudWatch, X-Ray, Splunk, New Relic dashboards and alerting

- Design platform APIs supporting REST, SOAP, GraphQL, OAuth2, streaming protocols, and A2A communication

Required Skills & Experience :

- 5+ years AWS platform engineering with Bedrock, AgentCore, EKS/ECR experience

- Expert Terraform IaC (modules, state management, AWS providers) and Jenkins pipeline development

- Deep gateway experience (Kong, MCP servers, LLM routing, rate limiting, guardrails)

- Strong observability implementation across CloudWatch/X-Ray/Splunk/New Relic stacks

- Enterprise identity patterns (IAM roles/SCPs, Cognito federation, Okta SSO)

- Distributed systems design for multi-tenant AI platforms with fault tolerance and compliance

Nice to Have :

- Previous enterprise AI platform builds (Bedrock Guardrails, AgentCore gateway deployments)

- Multi-model LLM gateway operations (model routing, cost optimization, SLAs)