Generative AI & Agentic AI Architect

Looking for a highly skilled GenAI and Agentic AI Architect with deep hands-on expertise across AWS AI/ML services, multi-agent systems, LLM orchestration, and cloud-native engineering.

This role centers on architecting and delivering scalable, secure, production-grade GenAI solutions using AWS AgentCore, AWS Bedrock, AWS Strands SDK, vector databases, and Python-based agentic AI solutions.

You will define reference architectures, code solutions, lead complex implementations, and drive adoption of advanced GenAI and agentic patterns across enterprise systems.

Key Responsibilities :

• Experience in pre-sales, solutioning, and deal shaping for large digital transformation programs.

GenAI & Agentic AI Architecture :

• Architect end-to-end GenAI-first and Agentic AI solutions on AWS using :

• Amazon Bedrock

• Amazon AgentCore

• Strands SDK Multi-Agents

• AWS Agents for Bedrock

• Knowledge Bases for Bedrock

• Guardrails for Bedrock

• Amazon SageMaker for model lifecycle & fine-tuning

• Design multi-agent systems supporting planning, reasoning, tool invocation, memory stores, and multi-step workflows.

• Develop high-performance RAG architectures using :

• Amazon OpenSearch Serverless

• Kendra

• Aurora + pgvector

• Redis Enterprise

• Implement secure prompt flows, tool integrations, embeddings pipelines, and evaluation frameworks optimized for AWS.

Application & Platform Engineering :

• Build Python-based microservices (FastAPI/Django) integrated with AWS AI stack.

• Design scalable architectures using AWS-native components :

• Lambda, Step Functions, EventBridge, SQS, DynamoDB, API Gateway, ECS/EKS

• Implement enterprise-class LLMOps using :

• Amazon SageMaker Pipelines

• AWS CodePipeline / CodeBuild

• CloudWatch / X-Ray observability

• Establish standards for :

• Agent governance & safety

• Secure model invocation

• Guardrails & content filters

• Latency & cost optimization patterns

Security, Reliability & Optimization :

• Ensure compliance with enterprise security standards using :