Logo
Medium

Platform AI Architect

Medium, Chicago, Illinois, United States, 60290

Save Job

Principal AI Architect will own the end-to-end architecture of the Agentic AI Platform across all layers including Meta-Agent, Orchestration, Agent Runtime, Memory/State, Tools, RAG, and Model layer. They will also define the technical blueprint, reusable patterns, AWS-native design, and multi-account strategy that power every agentinthe platform. Additionally, they will ensure the system is secure, scalable, well-governed, and aligned with enterprise standards for performance, reliability, and compliance.

Responsibilities

Architect the full Agentic Platform across AWS, Bedrock, LangChain/LangGraph, RAG, and Lakehouse components.

Define multi-step agent patterns (ReAct, planning, routing) and reusable architecture templates.

Own Bedrock model selection, embeddings strategies, knowledge bases, and guardrails design.

Design AWS foundations: VPC, Subnets, IAM, KMS, Secrets Manager, CloudTrail, Config.

Lead Lakehouse architecture: Glue, Athena, Lake Formation, S3 lakehouse patterns.

Enforce structured output schemas, prompt architecture, and cross-account separation.

Provide architectural sign-off for all agent workflows, tools, and platform extensions.

Ensure platform is scalable, observable, compliant, and optimized for cost and performance.

Qualifications

Deep AWS architecture expertise.

Proven experience building large-scale AI platforms.

Expert-level agent design skills.

Strong knowledge of AWS data stack.

Ability to drive platform-wide standards.

Demonstrated delivery in regulated industries.

$250,000 - $280,000 a year

#J-18808-Ljbffr