> For the complete documentation index, see [llms.txt](https://docs.cortex.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.cortex.io/improve/opex-review.md). # OpEx Review Agent {% hint style="info" %} OpEx Review Agent is in Research Preview and available to a limited group of customers. To request access, reach out to your Customer Success Manager. {% endhint %} The OpEx Review is an agent that helps engineering leaders run their OpEx program. It reads the signals across your organization, finds the operational root causes, and hands you a briefing at a regular cadence, so you can spend OpEx reviews digging in and making decisions. Every answer is traceable and grounded in your data. You can drill in immediately; ask it—in plain language—why a number moved, which teams are driving it, or how it compares to last quarter, and get to the root cause immediately. On the schedule you set, the agent analyzes your Cortex data, organized by the DRIVE framework (Delivery, Reliability, Initiatives, Vigilance, and Efficiency), identifies patterns and anomalies, and writes a plain-language briefing of what changed since the last cycle and what needs your attention.
OpEx Review in Cortex.
## Key features * **Scheduled, narrated reports** - The agent generates a report on the cadence you choose. The executive briefing explains what changed since the last cycle and flags areas of risk, in plain language. * **Fully configurable metrics, grouped by DRIVE pillar** - Cortex recommends a set of default metrics, but you decide what each pillar tracks. Choose from out-of-the-box [Eng Intelligence metrics](/improve/eng-intelligence.md), [custom metrics](/improve/eng-intelligence/custom-metrics.md), or [Scorecards](/standardize/scorecards.md), and add metrics only to the pillars you care about. * **AI insights on every metric** - Each metric includes its current value, the change from the prior period, its trend over time, and AI-generated insights that explain what's driving the movement and how it connects to other metrics. * **Bring your own dashboards** - Add dashboards you've already built in Cortex to OpEx Review, so you can pull in other analytics or visualizations you want to reference during your meetings. * **Dig deeper with AI chat** - [Open a chat](/improve/opex-review/using-opex-ai-chat.md) with Cortex AI about any metric; ask follow-up questions, dig into why a metric moved, and pull in related data without leaving the report. * **Shareable, point-in-time reports** - Each report the agent generates covers a fixed period. Use the report picker to move between periods, and share a link to the exact report you're viewing. ## Leveraging OpEx Review for your OpEx meetings An OpEx meeting is only as useful as the data powering it. Most of that data gets assembled manually, and it's expensive: hours per team, adding up to days of work across the org for a single review. Then during the meeting itself, teams review so many metrics that the signal gets lost, and real issues never make it into the conversation. OpEx Review handles the assembly for you and gives the meeting a single source of truth, so you can start with what's failing and who owns the fix instead of piecing health status together from disconnected tools. Expand the image below for tips on using OpEx Review during your meeting.
## Getting started with OpEx Review * [Configuring OpEx Review](/improve/opex-review/configuring-opex.md) - Tell the agent which metrics you want to track, set the report scope, and pick an analysis cadence. * [Using the OpEx Review report](/improve/opex-review/using-opex.md) - Review the agent's output, including the executive briefing, metrics broken down by team or domain, and customized dashboards. * [Using Cortex AI in OpEx Review](/improve/opex-review/using-opex-ai-chat.md) - Ask the agent follow-up questions about any metric without leaving the report.