Principal Engineer

A2Z; Denver (CO); On-site; USD 140,000 - 220,000; Full time

Job summary

A2Z in Denver, Colorado, is building a scalable, multi-tenant automotive AI platform. This role focuses on writing agent code, defining tool interfaces, and shipping production AI workflows on AWS Bedrock and MCP servers.

You will own the AI agent lifecycle end-to-end—from prompt design and tool use to guardrails, cost optimization, and production observability. You’ll establish the patterns the team uses to connect agents to production systems, evaluate output quality, and roll back when agents

Responsibilities

• Ownership and core development of agentic AI systems—design, build, and operate AI agent infrastructure on AWS Bedrock.
• End-to-end AI agent lifecycle—from prompts and tool design to guardrails, evaluation, cost optimization, and observability.

Job description

Why This Role Exists

We operate a multi-tenant automotive SaaS platform serving thousands of dealer groups across the United States. Our backend — event-driven serverless on AWS (Lambda, EventBridge, DynamoDB, S3, Step Functions) — orchestrates everything from dealer onboarding to inventory management to real-time transaction processing. That platform works. Now we need to make it think.

Scope & Scale

• 5000+ destination dealer tenants, each with isolated databases and per-tenant configuration
• Billions in annual Gross Merchandise Value (GMV) flowing through platform transactions
• Tens of thousands of API requests per minute across REST, SOAP, and event-driven integration surfaces
• Data pipelines spanning 6 integration domains with multi-protocol vendor connectivity

What You Will Own

• Ownership and core development of agentic AI systems — designing, building, and operating the AI agent infrastructure (AWS Bedrock, MCP servers) that powers intelligent automation across the platform. You are not advising on AI strategy; you are writing the agent code, defining the tool interfaces, building the evaluation harnesses, and shipping production AI workflows
• AI agent lifecycle end to end — from prompt engineering and tool-use design through guardrails, evaluation, cost optimization, and production observability. You own the patterns the team uses to build with AI: how agents connect to production systems, how we evaluate output quality, how we manage model costs at scale, and how we roll back when an agent misbehaves