PROJECT OVERVIEW
We are building an AI-powered media planning solution that transforms manual, expert-dependent workflows into an intelligent conversational experience. The platform uses a multi-agent architecture orchestrated by AWS Bedrock to guide users through complex campaign planning, integrating real-time data from multiple enterprise systems, and generating mathematically optimized media proposals. The solution combines agentic AI, knowledge retrieval (RAG), mathematical optimization engines, and enterprise integrations to deliver end-to-end workflow automation from opportunity discovery to campaign activation.
ROLE OVERVIEW
As an AI/ML Engineer, you will design and implement the agentic core of the platform, building specialized agents that collaborate to complete complex planning workflows. You will architect the agent orchestration logic, implement tool integrations via Model Context Protocol (MCP), configure AWS Bedrock components (Agents, Knowledge Bases, Guardrails), and establish observability infrastructure to ensure transparency and debuggability. This role requires deep expertise in AWS generative AI services, prompt engineering, agentic systems, and production ML operations.
MUST HAVE
- Advanced/Fluent English skills
- Experience building production ML/AI systems on AWS
- Hands-on expertise with AWS Bedrock Agents (AgentCore), including agent orchestration, tool integration, and multi-agent coordination
- Experience with AWS Bedrock Knowledge Bases (RAG), OpenSearch Serverless, and vector embeddings
- Proficiency in AWS Bedrock Guardrails for content filtering, PII protection, and policy enforcement
- Experience implementing agentic workflows with tool calling, function invocation, and state management
- Experience with prompt engineering, temperature tuning, and LLM behavior optimization
- Strong understanding of AWS observability tools: CloudWatch, X-Ray, and distributed tracing
- Experience with API Gateway, DynamoDB, and event-driven architectures
- Familiarity with Infrastructure as Code (Terraform or CDK)
- Strong understanding of RESTful APIs and integration patterns
- Experience with CI/CD pipelines for ML systems
- Ability to design error handling, retry logic, and graceful degradation strategies
- Strong documentation and communication skills
NICE TO HAVE
- Experience with Model Context Protocol (MCP) or similar tool integration frameworks
- Knowledge of Amazon Cognito for authentication and SSO integration
- Experience with CloudFront, S3, and WAF for secure frontend delivery
- Familiarity with conversational AI design patterns (Smart Cards, hybrid input modalities)
- Knowledge of session management and conversation state persistence
- Understanding of compliance requirements (PII handling, audit trails, data retention)
- Experience with AWS Cost Optimization for Bedrock and OpenSearch workloads
- Familiarity with LLM evaluation frameworks and agent performance metrics
- Experience with multi-turn dialogue management and context preservation
- Knowledge of agent reasoning visualization and explainability techniques
- Experience with AWS Step Functions for complex workflow orchestration
If you like it, just apply and good luck!
#LI-JM2