Rescale Builds AI Triage Agent with AWS Strands Agents

Disclosure: this post contains affiliate links. If you buy through them we may earn a commission at no extra cost to you. Learn more.

TL;DR: Rescale built an AI triage agent using AWS's AI-Driven Development Lifecycle (AI-DLC) framework, going from problem definition to staging deployment in five days and production within weeks. The agent automates classification and routing of nightly end-to-end test failures, running on Amazon EKS with Claude Opus 4.8 on Amazon Bedrock.

What's New

Rescale, a digital engineering platform for the AI era, has successfully deployed an intelligent test-triage agent using AWS's AI-Driven Development Lifecycle (AI-DLC) framework. The company went from initial problem definition to staging deployment in just five days, with full production deployment completed within weeks.

The triage agent addresses a critical operational challenge: classifying and routing automated end-to-end test failures. Previously, engineers spent significant time manually reviewing test reports, examining logs from multiple systems, and coordinating across teams to resolve failures that blocked production releases.

The solution runs as a step within Rescale's existing Argo Workflows CI/CD pipeline on Amazon Elastic Kubernetes Service (Amazon EKS). When nightly Playwright test suites produce failures, the agent automatically invokes Claude Opus 4.8 on Amazon Bedrock to classify each failure into categories including infrastructure issues, code bugs, flaky tests, or environment problems. It retrieves relevant historical failure patterns from a knowledge base, queries pod logs through Model Context Protocol (MCP) for runtime evidence, and generates structured remediation reports with specific file references, function names, and recommended fixes.

The AI-DLC framework emphasizes a collaborative development approach where "AI executes, humans decide." AI generates work plans and code but surfaces decision points for human validation rather than making assumptions. The methodology uses focused group sessions called Mob Elaboration for defining requirements and Mob Construction for building collaboratively.

Pricing and Availability

The article does not provide specific pricing information for the tools and services used. The AI-DLC framework is described as open source. Pricing for Amazon Bedrock, Amazon EKS, and other AWS services would apply according to standard AWS pricing models.

How to Try It

Rescale's implementation followed a structured five-day in-person workshop with AWS architects. The engagement included an inception phase (days 1-2) for scoping requirements and a construction phase (days 2-4) for architecture and implementation.

The team started with a vision document and technical environment document, then worked through AI-generated clarification questions to produce scoped requirements. Five user stories were consolidated into four units of work: a knowledge base for historical failure patterns, the AI triage agent as the core classification engine, Argo Workflow integration for deployment, and an evaluation harness for accuracy measurement.

Organizations interested in similar implementations would work with AWS to structure comparable workshops using the AI-DLC methodology. The approach is designed to be extended across engineering organizations for other use cases beyond test triage.

Who It's For

This solution is designed for engineering teams facing operational overhead from repetitive but critical tasks like triaging test failures, particularly those running distributed services across multiple repositories. It's especially relevant for organizations where automated tests gate production releases and unresolved failures require significant coordination across teams.

Rescale serves R&D and engineering teams across aerospace, automotive, energy, life sciences, semiconductor, and manufacturing sectors. The AI-DLC framework is applicable to teams looking to incorporate AI assistance throughout the engineering lifecycle while maintaining human accountability for results.

Alternatives

While this article focuses on AWS's specific implementation, other DevOps automation and AI-powered testing platforms exist in the market. Make offers workflow automation capabilities that could complement CI/CD pipelines, though with different technical approaches than the agent-based system described here. Organizations might also consider building custom solutions using other cloud providers' AI services or open-source LLM frameworks, though these would lack the structured AI-DLC methodology AWS provides.