This is the Trace Id: 0767695a45c389d73a0db74ec90f4bd4
10/28/2026

InEight cuts incident investigation effort 60% with Azure SRE Agent

More than 850 companies manage over $1 trillion in project work on InEight software. Supporting it means a global Azure estate where one defect can surface across hundreds of databases—and a typical incident took two hours to diagnose.

InEight chose Azure SRE Agent to do that tracing for them—reading their own telemetry, code, and deployment history as one picture. Six layers of governance keep it from touching production without an engineer’s approval.

Incident investigation and initial bug investigation are each down 60%, and one on-call engineer can now handle roughly 80% of incidents without waking anyone else.

InEight Technology

Manage reliability at global scale

InEight provides project controls and project management software for organizations delivering complex capital projects. More than 850 companies use its 13 integrated products to manage over $1 trillion in project work across construction, engineering, transportation, mining, energy, and other asset-intensive industries.

Supporting that business means operating more than 72,000 Microsoft Azure resources and 34,000 SQL databases across a global estate spanning more than 20 Azure regions and generating roughly two terabytes of telemetry each day. On the reliability side, teams see roughly 40 to 50 alerts per product across InEight’s 13-product suite. A single defect can surface across hundreds of databases, making investigation a problem of both volume and complexity.

Before Azure SRE Agent, an investigation often began as what Jim Ellerbeck, VP of Engineering, DevOps and R&D at InEight, called a “scavenger hunt.” Engineers moved among SQL diagnostics, Log Analytics, Application Insights, message queue telemetry, deployment records, source repositories, and internal documentation—sometimes checking 10 different sources just to establish a timeline.

“It was always a nightmare to have a single pane of glass view of all those thousands of resources,” says Babji Talluri, Senior DevOps Engineer at InEight. “We had to go into 10 different sources to pull the information together.”

The problem was not a shortage of monitoring. InEight already had thousands of alerting rules. The problem was turning fragmented signals into a coherent explanation. With 13 products and globally distributed deployment stamps, engineers needed to know whether an issue was isolated or systemic, which customers and services were affected, what had changed, and whether the answer was an infrastructure adjustment or a code fix.

Babji Talluri, Senior DevOps Engineer, InEight

“It was always a nightmare to have a single pane of glass view of all those thousands of resources. We had to go into 10 different sources to pull the information together.”

Babji Talluri, Senior DevOps Engineer, InEight

Move from alert to root cause without pulling in every specialist

InEight adopted Azure SRE Agent to change that operating model. The agent brings those fragmented signals into a live reasoning loop grounded in InEight’s own environment: telemetry tied to the relevant product, deployment, source code, and prior operational knowledge, so teams can move from an alert to a probable root cause and a mitigation path without rebuilding context by hand.

Before Azure SRE Agent, a typical incident took about two hours to detect and diagnose; more complex cases took three to four hours to assemble context, review telemetry, determine a short-term mitigation, and reach a code-level conclusion. That work often pulled database, application, security, product, and QA specialists into the same investigation—even overnight.

“It used to take anywhere between three to four hours to establish the context, review the telemetry, and come up with a meaningful short-term and long-term mitigation—and we might still end up reaching out to the application teams overnight and waking them up,” Talluri explains. “That strain and toil has been reduced.”

InEight now reports a 60% reduction in incident investigation effort, with the agent producing the diagnosis, remediation guidance, and root-cause context along the way. Talluri estimates that one on-call engineer can now resolve about 80% of incidents without waking specialists or pulling the whole team together.

Babji Talluri, Senior DevOps Engineer, InEight

“It used to take anywhere between three to four hours to establish the context, review the telemetry, and come up with a meaningful short-term and long-term mitigation—and we might still end up reaching out to the application teams overnight and waking them up. That strain and toil has been reduced.”

Babji Talluri, Senior DevOps Engineer, InEight

Turn faster diagnosis into engineering capacity

The gains extend into software engineering. Initial bug investigation time has also fallen 60% because the agent connects production evidence to the relevant code path, packages the findings for a developer, and moves the work into Azure DevOps.

The team has applied that approach to multiple complex, code-aware root cause analyses, pipeline failure investigations, and performance troubleshooting scenarios across build and release workflows. Those investigations were validated in sandbox environments, and some of the resulting improvements are already in production.

When a failure pattern is confirmed in one production stamp, InEight can look for the same condition across other stamps, customers, and products. Reliability improvements no longer stop at the boundary of a single incident.

“The real advantage of AI isn’t that it can write code fast,” Ellerbeck notes. “The real advantage is finding the problem faster. Reducing the investigation time accelerates every downstream engineering activity.”

InEight’s goal is to reduce engineering hours spent on this operational work by 50% across a DevOps and DBA organization of roughly 24 to 30 people—capacity that can move back into product quality, customer-facing features, and the company’s next wave of AI innovation.

Jim Ellerbeck, VP of Technology - DevOps, R&D, InEight

“The real advantage of AI isn’t that it can write code fast. The real advantage is finding the problem faster. Reducing the investigation time accelerates every downstream engineering activity.”

Jim Ellerbeck, VP of Technology - DevOps, R&D, InEight

Govern an operational platform at enterprise scale

As adoption expanded, Azure SRE Agent moved beyond a troubleshooting interface to an operational backbone. It works across an environment encompassing roughly 100 Azure resource types, connecting operational evidence with the code, data, pipelines, and Azure DevOps workflows engineers already use—creating work items and pull requests, authoring pipeline files, triggering build and release pipelines, summarizing security findings, and tying investigations back to test cases and deployment evidence.

That reach is bounded by deliberate governance. InEight built six layers of controls around the platform, including a version-controlled rule set of 46 guardrails, root-level instructions, hooks, role separation, managed identity, and always-on review mode. Production agents are constrained by geography and permission; potentially disruptive operations such as scaling, restarting, deleting, or changing SQL resources are blocked, and engineers retain approval over actions. As Ellerbeck describes it, the agent spins fast but stays contained inside the governance model InEight already had in place. This is a practical model for enterprise adoption: preserve proven DevOps controls, encode them into the agent’s operating environment, and expand only as the organization builds confidence.

Turn expert knowledge into a reusable asset

The operating knowledge that mattered most lived with a handful of senior specialists. InEight spent months capturing operational practices, skills, and engineering knowledge before expanding into production; its repository now holds roughly 270 documents, putting knowledge that once depended on individual experts into a governed, version-controlled resource any on-call engineer can draw from.

“The expert knowledge sat with a handful of senior engineers,” Talluri adds. “We are institutionalizing and preserving all that knowledge and skill into a centralized repository that is version-control gated.”

The improvement loop is operational, not theoretical. A validated fix discovered for one product can be checked across other products and customer environments. Investigation findings become work items, pull requests, tests, and pipeline checks. Daily reports compare what changed, what remains unresolved, and where attention is needed next. Each cycle makes the platform more grounded in how InEight actually builds and runs software.

Babji Talluri, Senior DevOps Engineer, InEight

“The expert knowledge sat with a handful of senior engineers. We are institutionalizing and preserving all that knowledge and skill into a centralized repository that is version-control gated.”

Babji Talluri, Senior DevOps Engineer, InEight

Move from incident response to engineering intelligence

InEight’s next phase is to make Azure SRE Agent part of everyday engineering leadership. Agent-generated views of Azure DevOps data already support capacity planning, quality programs, release decisions, and roadmap discussions, and parts of InEight’s health and security reviews, software development lifecycle (SDLC) workflows, bug and incident triage, and pipeline automation run on the platform today. Next, InEight plans to broaden automation across bug triage, incident response, and SDLC processes, and to extend it into security and vulnerability management, build and release certification, code reviews, test case generation, code refactoring and optimization, and load test analysis and triage. Routine, repetitive, evidence-heavy work moves to the platform, while people remain responsible for judgment, approvals, architecture, and customer outcomes.

“SRE Agent is helping InEight transform how engineering operates,” Ellerbeck says. “By embedding AI into software delivery, reliability engineering, quality assurance, security, and operational workflows, we are reducing manual effort, accelerating delivery, improving stability, and creating a scalable foundation for future growth. Incident investigation is down 60 percent, build failure triage down 60 percent, and bug investigation down 60 percent. With Azure SRE Agent, we can reason across source code, live telemetry, and Azure infrastructure simultaneously—in a single conversation. For a company operating a large suite of integrated products on Azure, that capability is not incremental. It is transformational.”

In four months, the team moved from an initial pilot to production use cases spanning incident response, bug triage, security, delivery, and planning. For Ellerbeck, the point of all of it is where engineering attention goes toward what InEight wants to build next, rather than repeatedly troubleshooting what it already has.

Learn more about Azure SRE Agent.

Discover more about InEight on LinkedIn.

Jim Ellerbeck, VP of Technology - DevOps, R&D, InEight

“With Azure SRE Agent, we can reason across source code, live telemetry, and Azure infrastructure simultaneously—in a single conversation. For a company operating a large suite of integrated products on Azure, that capability is not incremental. It is transformational.”

Jim Ellerbeck, VP of Technology - DevOps, R&D, InEight

Take the next step

Fuel innovation with Microsoft

Explore more customer stories

Find out how customers are achieving more with Microsoft products and solutions.
A man wearing headphones and smiling.

Talk to an expert about custom solutions

Let us help you create customized solutions and achieve your unique business goals.
Three people in a meeting room.

Transform work with Microsoft AI

Bring intelligence into the flow of work and help your organization achieve its goals with secure, scalable AI solutions.

Follow Microsoft