AWS Agentic AI Architect

Applogika; Charlotte (NC); Hybrid; USD 180,000 - 240,000; Full time

Job summary

Applogika is hiring an Agentic AI Architect to design and hands-on build autonomous agent systems on AWS, including Bedrock AgentCore, with a strong focus on production delivery. Expect roughly 60% hands-on building and 40% design, advisory, and leading delivery across customer projects.

Ideal candidates have 10+ years in software or solutions architecture, 4+ years on AWS, and 2+ years shipping generative AI to production.

Qualifications

• 10+ years in software engineering or solutions architecture
• 4+ years on AWS
• 2+ years shipping generative AI to production, including agentic systems
• Hands-on experience with Bedrock and Claude models
• Experience with agent frameworks (Strands Agents, LangGraph, or CrewAI) and MCP
• RAG and vector store work with measurable retrieval improvements
• Strong Python and TypeScript coding
• Daily use of AI coding agents like Claude Code
• Clear communication with engineers and executives

Responsibilities

• Architect single- and multi-agent systems on Amazon Bedrock AgentCore (Runtime, Memory, Gateway, Identity, Observability)
• Write production Python and TypeScript for agents, tools, and MCP servers
• Build RAG pipelines end to end: ingestion, chunking, embeddings, hybrid retrieval, reranking, and citations
• Stand up and tune vector stores: OpenSearch Serverless, Aurora pgvector, S3 Vectors, Bedrock Knowledge Bases
• Build eval harnesses for accuracy, groundedness, tool-call correctness, latency, and cost per task
• Apply guardrails, human-in-the-loop checkpoints, and AWS Well-Architected security to agents
• Run Working Backwards/discovery workshops with customers and take proofs of concept to production
• Mentor engineers through our FDE program

Skills

Python proficiency

TypeScript

AWS expertise

Agent frameworks knowledge

Communication

Tools

Amazon Bedrock

Anthropic Claude

Strands Agents

LangGraph

CrewAI

MCP

OpenSearch Serverless

Aurora pgvector

S3 Vectors

CDK

Terraform

Claude Code

Job description

Agentic AI Architect - Applogika | Charlotte, NC (Hybrid) or Remote, US | Full-time

About Applogika

Applogika is an AWS Advanced Tier Services Partner and Anthropic Claude Select-Tier Partner, We hold 8 AWS Competencies, including AI Services and AI Software, and we deliver generative and agentic AI on Amazon Bedrock and Claude for customers on AWS cloud.

The role

We're hiring an Agentic AI Architect who designs autonomous agent systems on AWS and then builds them, hands on keyboard, through to production. This is not a slideware role. Expect about 60% building and 40% designing, advising, and leading delivery.

What you'll do

• Architect single- and multi-agent systems on Amazon Bedrock AgentCore (Runtime, Memory, Gateway, Identity, Observability)
• Write production Python and TypeScript for agents, tools, and MCP servers, using Claude Code as your daily driver
• Build RAG pipelines end to end: ingestion, chunking, embeddings, hybrid retrieval, reranking, and citations
• Stand up and tune vector stores: OpenSearch Serverless, Aurora pgvector, S3 Vectors, Bedrock Knowledge Bases
• Build eval harnesses for accuracy, groundedness, tool-call correctness, latency, and cost per task
• Apply guardrails, human-in-the-loop checkpoints, and AWS Well-Architected security to agents that act in real systems
• Run Working Backwards/discovery workshops with customers and take proofs of concept to production
• Mentor engineers through our FDE program

What you bring

• 10+ years in software engineering or solutions architecture, 4+ on AWS
• 2+ years shipping generative AI to production, including at least one agentic system that takes real actions
• Hands-on experience with Amazon Bedrock and Anthropic Claude models (tool use, prompt engineering, model selection)
• Experience with an agent framework such as Strands Agents, LangGraph, or CrewAI, plus MCP
• RAG and vector store work with measurable retrieval improvements
• Strong Python, working TypeScript, and CDK or Terraform
• Daily use of AI coding agents like Claude Code
• Clear communication with engineers and executives