{"version":1,"type":"story","url":"https://digestai.news/story/cornerstone-ondemand-cuts-database-diagnosis-time-by-78-with-amazon-be","json":"https://digestai.news/story/cornerstone-ondemand-cuts-database-diagnosis-time-by-78-with-amazon-be.json","markdown":"https://digestai.news/story/cornerstone-ondemand-cuts-database-diagnosis-time-by-78-with-amazon-be.md","slug":"cornerstone-ondemand-cuts-database-diagnosis-time-by-78-with-amazon-be","headline":"Cornerstone OnDemand cuts database diagnosis time by 78% with Amazon Bedrock","summary":"Cornerstone OnDemand, a workforce solutions provider with 140 million users, built **Orion AI**, a multi-agent system using **Amazon Bedrock** and **Strands Agents** to automate database operations. Before Orion AI, its **Enterprise DataOps team** spent up to **45 minutes per incident** manually diagnosing issues across tools and logs. The new system reduced diagnosis time to **10 minutes**, a **78% improvement**, while also automating **10+ manual steps** in database lifecycle workflows and cutting alert noise by **65%**. A **three-person team** developed Orion AI in **six months**, integrating specialized agents for diagnostics, lifecycle management, and real-time monitoring, all while maintaining data privacy through AWS controls.\n\nOrion AI works by routing requests to domain-specific agents—each handling a narrow function like SQL diagnostics or blocking analysis—using **keyword-based routing** for speed and **semantic search** for ambiguous queries. The system bypasses conversational memory for live metrics to ensure accuracy. Deployed on **Amazon ECS**, Orion AI also includes guardrails for safety, observability via **CloudWatch and X-Ray**, and cost tracking. Cornerstone shares its design principles—**domain-splitting agents, hybrid routing, and session-scoped memory**—as reusable patterns for other operations teams.","keyPoints":["Cornerstone OnDemand’s Orion AI slashed database diagnosis time from 45 to 10 minutes, a 78% reduction","System automates 10+ manual steps in database workflows and reduces alerts by 65% via deduplication","Built with Amazon Bedrock and Strands Agents, deployed on AWS ECS in six months by a three-person team"],"whyItMatters":"Orion AI demonstrates how **multi-agent systems** can transform reactive IT operations into proactive, automated workflows, cutting manual labor and improving real-time diagnostics. Its modular design and AWS integration offer a blueprint for other enterprises to adopt similar efficiency gains in database and DevOps tasks.","category":{"slug":"enterprise","name":"Enterprise & Industry","url":"https://digestai.news/category/enterprise"},"entities":{"companies":["Cornerstone OnDemand","Amazon","AWS"],"models":["Amazon Bedrock","Amazon Titan Text Embeddings V2","Amazon Nova 2 Lite"],"people":[]},"firstPublishedAt":"2026-10-07T15:38:28Z","updatedAt":"2026-10-07T15:38:28Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"AWS Machine Learning Blog","title":"How Cornerstone OnDemand cut database diagnosis by 78% with Amazon Bedrock","url":"https://aws.amazon.com/blogs/machine-learning/how-cornerstone-ondemand-cut-database-diagnosis-by-78-with-amazon-bedrock","publishedAt":"2026-10-07T15:38:28Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Cornerstone OnDemand cuts database diagnosis time by 78% with Amazon Bedrock\", 7 October 2026, https://digestai.news/story/cornerstone-ondemand-cuts-database-diagnosis-time-by-78-with-amazon-be","publisher":"Digest AI","title":"Cornerstone OnDemand cuts database diagnosis time by 78% with Amazon Bedrock","datePublished":"2026-10-07T15:38:28Z","url":"https://digestai.news/story/cornerstone-ondemand-cuts-database-diagnosis-time-by-78-with-amazon-be"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}