Operations Agent: 3 Minute Demo

Let's take a peek at how data and AI reliability teams might work in the future. In this scenario, I want to learn about a data asset and make sure the monitoring and alert strategy are on point. Let's start by wondering why I should even care about this table in the first place. We can see here our F sales table powers some executive reporting as well as our five regional sales attainment dashboards. Good to know. Now, let's see what monitors I should add if I want to maximize my coverage and catch issues before my colleagues in sales start yelling. Here, I'm asking the operations agent to recommend some monitors for me. This enables different members of the organization to participate in monitoring even if they aren't monitoring experts. There are 10 monitors currently deployed. This is a fairly well- covered table, but there are some gaps. One gap that stands out to me is that there are key fields with no monitors to alert me to null values or otherwise data completeness issues. Let's go ahead and fix that. Let's create a metric monitor for our null percentage on all applicable fields within this table. And with one easy request, we can see that Monte Carlo's operation agent has defined the appropriate monitor in YAML using monitors as code. Now, this can be easily deployed as part of my CI/CD process. We can see all relevant fields like software ID, developer ID, opportunity ID. All the relevant fields within this table have been covered as part of this metric monitor. Now, let's take a look to see if there are any monitors on this table that do not have corresponding audiences. In other words, alerts that might be firing, but no one's at home listening. This is a key triage workflow, and we can see we have a few monitors that aren't sending alerts anywhere. Some of them have been paused, but some are not only active, but have breached conditions. Let's figure out and see where they should go. Here we'll ask the operations agent to make recommendations for us based on its knowledge of our lineage as well as owners and notification audiences. The recommendations here make sense and save me a ton of time, likely hours if I was trying to figure this out using more manual workflows. Finally, let's dive deeper into an active alert and learn more about this breach for our invoice quantity enforcement monitor. As the operations agent provides us information on this alert, let's review what we've covered in this video. We were able to, as a non-technical member of the team, learn more about key assets within our environment, ensure they were monitored correctly and send alerts to the right audience. We're able to do this using natural language in just a few questions rather than have to gather this information across multiple different systems or even across the Monte Carlo UI. If you're interested in supercharging your team with AI powered workflows, you need Monte Carlo in the operations agent. Be sure to request a demo