Rescale built an AI triage agent in 5 days with Strands Agents and Claude on Bedrock
AWS ML Blog · rss · 2026-10-23
An AWS ML Blog deep-dive on how Rescale used the AI-DLC methodology and Strands Agents to build an agent that triages nightly end-to-end test failures — problem definition to staging in 5 days, production within weeks.
How it works: the agent triggers from the nightly Playwright suite, invokes Claude Opus 4.8 on Amazon Bedrock to classify each failure (infrastructure / code bug / flaky test / environment), pulls historical failure patterns from a knowledge base, fetches pod logs via MCP, and emits a structured remediation report with file references for engineer review.
AI-DLC highlights:
- "AI executes, humans decide": humans set scope boundaries and approve each plan
- Mob Elaboration / Mob Construction sessions replace async handoffs; the team answered 20+ AI clarification questions live
- Two architecture pivots mid-build: Bedrock AgentCore → existing EKS/Argo cluster; third-party logging → CloudWatch with the MCP pattern unchanged
Notable design choices include conclusions-over-raw-rows tool outputs and human approval for all production changes.
More from coding & agent
- Developers Don't Code Anymore — Now Their Job Is Rotating Secrets Daily — Dan_Jeffries1 · 2026-10-23
- New paper: coding agents in a shared sandbox self-create rules, reciprocity debts, and coalitions — nandofioretto · 2026-10-23
- Developer runs AI project managers that manage their own AI agents to build his game — draginol · 2026-10-23
- RentAKaren: a real service that rents humans to handle complaints your AI agent can't win — mathemagic1an · 2026-10-23
- Clerk's Auth for MCP growing 13% week-over-week, fastest since 2023 — stuffyokodraws · 2026-10-23
- Shopify builds remote mobile dev environments for AI agents — rseroter · 2026-10-23