> ## Content Index > Fetch the complete content index at: https://www.stirup.si/llms.txt > Use this file to discover other available public pages before exploring further. # Rescale Deploys AI Triage Agent to Automate Test Failure Classification - URL: https://www.stirup.si/rescale-deploys-ai-triage-agent-to-automate-test-failure-classificatio-b29f5da2/ - Published: 2026-10-28T16:58:11.000Z - Updated: 2026-10-28T16:58:11.000Z - Description: Rescale implemented an AI triage agent using Strands Agents and AI-DLC on AWS, improving classification agreement with human triage from 44% to 68% and reducing time blocking production deployment by 20%. - Author: Satish Das - Tags: Cloud, News, Stirup Research ## Problem Definition and Triage Agent Overview Rescale, a digital engineering platform, faced significant operational overhead due to engineers manually classifying automated test failures. Prior to the deployment of the AI triage agent, engineers spent considerable time on repetitive tasks related to classifying and routing test failures, which hindered their ability to focus on more complex engineering challenges. The triage agent was a well-scoped problem with a tight timeline, prompting Rescale to explore solutions that could streamline this process. By leveraging Strands Agents and AI-DLC, Rescale aimed to automate the classification and routing of test failures, reducing manual intervention and allowing engineers to concentrate on higher-value tasks. This approach highlights the potential of AI-driven tools in addressing specific operational inefficiencies within engineering workflows. ## Implementation Details and Architecture Rescale’s AI triage agent is implemented using the Strands SDK, enabling concurrent analysis of nightly end-to-end test failures with structured output in JSON and HTML. The agent is containerized and deployed on Amazon Elastic Kubernetes Service (Amazon EKS), as noted in the solution overview and integration details. It is integrated into Rescale’s existing Argo Workflows CI/CD pipeline on Amazon EKS, allowing seamless automation of failure classification and routing. This architecture supports scalability and efficient resource utilization, ensuring that test failures are processed and prioritized effectively without disrupting the development workflow. The use of Strands Agents and AI-DLC technologies facilitates the agent’s ability to handle complex failure patterns and provide actionable insights for engineering teams. ## Performance Metrics and Outcomes Rescale implemented an AI triage agent using the AI-Driven Development Lifecycle (AI-DLC) framework on AWS, which improved classification agreement with human triage from 44% to 68% as reported by [welcome.ai](https://welcome.ai/content/rescales-ai-triage-agent-enhances-efficiency-and-accuracy-on-aws?ref=stirup.si). This enhancement was achieved through the integration of Strands Agents, which allowed for more accurate issue classification and prioritization. According to [AWS Machine Learning](https://aws.amazon.com/blogs/machine-learning/how-rescale-built-an-ai-triage-agent-with-strands-agents-and-ai-dlc/?ref=stirup.si), the agent’s classification agreement with human triage has risen from 44% at initial deployment to 68% today. Additionally, [welcome.ai](https://welcome.ai/content/rescales-ai-triage-agent-enhances-efficiency-and-accuracy-on-aws?ref=stirup.si) notes that the automation has led to a 20% decrease in time blocking production deployment while test failures are analyzed, allowing for faster releases of new changes. These improvements highlight the effectiveness of combining AI-DLC with Strands Agents in streamlining the deployment process and enhancing overall system reliability. ## Operational Integration and Evaluation During the operational integration and evaluation phase, Rescale implemented a PyTest-based evaluation framework with live and stubbed test modes, establishing an initial classification-agreement baseline to measure the AI triage agent's performance ([AWS Machine Learning](https://aws.amazon.com/blogs/machine-learning/how-rescale-built-an-ai-triage-agent-with-strands-agents-and-ai-dlc/?ref=stirup.si)). The agent was successfully demoed to Rescale’s CTO and VP of Engineering, accurately classifying an infrastructure failure and a real code bug, including specific file-level fix recommendations ([AWS Machine Learning](https://aws.amazon.com/blogs/machine-learning/how-rescale-built-an-ai-triage-agent-with-strands-agents-and-ai-dlc/?ref=stirup.si)). The system utilizes Claude Opus 4.8 on Amazon Bedrock to classify failures, retrieve historical failure patterns, query logs for evidence, and generate structured remediation reports for engineers to review ([welcome.ai](https://welcome.ai/content/rescales-ai-triage-agent-enhances-efficiency-and-accuracy-on-aws?ref=stirup.si)). These steps ensured the agent could be effectively evaluated and integrated into Rescale’s operational workflows. ## Broader Implications and Future Plans The broader implications of Rescale's AI triage agent extend beyond immediate operational efficiencies. As noted in [AWS Machine Learning](https://aws.amazon.com/blogs/machine-learning/how-rescale-built-an-ai-triage-agent-with-strands-agents-and-ai-dlc/?ref=stirup.si), the triage agent is part of a larger initiative to deliver digital engineering capabilities tailored for the AI era. This signals a strategic shift toward integrating AI into core engineering workflows. According to [welcome.ai](https://welcome.ai/content/rescales-ai-triage-agent-enhances-efficiency-and-accuracy-on-aws?ref=stirup.si), Rescale plans to establish an AI-DLC Center of Excellence, suggesting a long-term commitment to AI-native engineering practices. The triage agent itself, as described in [welcome.ai](https://welcome.ai/content/rescales-ai-triage-agent-enhances-efficiency-and-accuracy-on-aws?ref=stirup.si), automates the classification and routing of test failures, significantly reducing the manual effort required by engineers. These developments highlight Rescale's intent to position itself as a leader in AI-driven digital engineering, though specific outcomes or timelines for these initiatives remain unquantified. ## Sources and reporting notes - [AWS Machine Learning](https://aws.amazon.com/blogs/machine-learning/how-rescale-built-an-ai-triage-agent-with-strands-agents-and-ai-dlc/?ref=stirup.si) - [welcome.ai](https://welcome.ai/content/rescales-ai-triage-agent-enhances-efficiency-and-accuracy-on-aws?ref=stirup.si) - [agentscookbook.com](https://agentscookbook.com/?ref=stirup.si) - [stellarmaterialsolutions.com](https://stellarmaterialsolutions.com/ai-briefing/?ref=stirup.si) Based on published source material. No independent hands-on testing is claimed.