[Job - 30521] AI Engineer Senior (Cloude Code)

AI Engineer Sr

Project Overview

We are building an enterprise-grade agentic AI platform for a leading media and broadcasting technology company that serves thousands of radio stations across multiple cloud products. This serverless, multi-tenant architecture leverages Amazon Bedrock AgentCore ecosystem to orchestrate intelligent automation across several SaaS platforms, enabling conversational AI interfaces for operations management, knowledge retrieval, data analysis, and system mutations. The platform is designed to handle 50,000+ daily interactions while maintaining strict tenant isolation, security guardrails, and operational observability.

The technical implementation focuses on hybrid knowledge base integration, Model Context Protocol (MCP) gateway development, agentic skill orchestration using the Strands Python framework, and implementing dual-layer security architecture with human-in-the-loop governance for high-impact operational changes. This is a greenfield opportunity to shape the future of AI-powered operations in the broadcasting industry.

Must Have Experiences

AWS Bedrock & AgentCore components (Runtime, Gateway, Memory, Identity, Observability)

Python development with AI/ML frameworks (Strands, LangChain, or similar)

Vector databases & RAG architectures (OpenSearch, Pinecone, or similar)

Model Context Protocol (MCP) & API gateway integration

Advanced English level

Agentic SDLC practices (prompt engineering, evaluation, HITL workflows)

AWS serverless services (Lambda, Step Functions, EventBridge, S3, CloudWatch)

Nice to Have Experiences

LLM security & guardrails implementation (prompt injection prevention, content filtering)

Multi-tenancy & identity management in SaaS environments

Intent classification & conversation flow management

Observability & performance monitoring (distributed tracing, token analytics)

LLM-as-a-Judge evaluation patterns & quality assurance pipelines

Event-driven architectures & real-time telemetry processing

Foundation model selection, fine-tuning & prompt optimization