Information technology / Enterprise software
Microsoft: Running the Smart DRI Agent, which gathers incident information and guides troubleshooting for on-call engineers (DRIs)
- Company
- Microsoft
- Country
- United States
- Adoption stage
- In operation
- Source published
- Date basis
- The date the source was published. It can differ from the date adoption started.
- How the source was checked
- Read the full source text
The work problem
According to an article published on October 15, 2026 by Microsoft's internal IT organization, engineers in Microsoft Digital take turns as designated responsible individuals (DRIs) and handle time-sensitive work such as investigating incidents, monitoring service health and reviewing security work. The information they needed was spread across several portals and dashboards, troubleshooting guides, SharePoint sites and OneNote notebooks, and a DRI often had to elevate production permissions repeatedly just to see an incident. Engineers new to a service took even longer because they had to find the right subject matter expert.
Technology and data
Engineers from Microsoft Digital and Microsoft HR turned the Smart DRI Agent, which began at an internal hackathon, into a tool used inside Microsoft Teams. Microsoft Semantic Kernel chooses which functions to call for each request across domains such as incident management and Service 360, and past incidents are indexed in Azure AI Search with added custom logic for judging similarity so that useful similar cases are returned. When an incident comes in, the agent pulls the traces, logs and exceptions from the last few minutes of telemetry, searches the troubleshooting guides and presents suggested steps and historical context, and it also sends a morning summary of what an engineer needs to complete that day. Human DRIs remain responsible for operational decisions.
Results
According to the article, the Smart DRI Agent now supports more than 40 teams at Microsoft and has delivered more than 7,800 AI-assisted interactions and nearly 6,000 proactive incident notifications. Microsoft estimated that this has saved about 600 engineering hours per month. Bhupesh Guptha Muthiyalu, a principal engineering manager in Microsoft HR, said DRIs have ready-made information in hand and are typically able to resolve incidents quickly.
Limits and open questions
The 600 hours is Microsoft's estimate, and the article does not say how much incident resolution time fell.
Sources
- Automating engineering operations at Microsoft with a Smart DRI Agentmicrosoft.com, Accessed
Compiled from public sources. These are not results from ATF Works customers.