In this lesson we cover:
-
Introduction
-
Accessing the Dashboard
-
What the Dashboard Shows
-
What Counts as an AI Call
-
Monitoring Agents
-
Find your early adopters, and put them to work.
-
Spot who hasn't started yet, and check in early.
-
Watch for unusually heavy usage too.
-
-
Automation Allowance & Forecast
Introduction
The dashboard is where Agent usage is tracked, and you'll use it slightly differently depending on your permission levels.
Most users will see just their own usage activity reported on in the Dashboard.
However, users with the right permissions will have a holistic view of all users activity. Additionally, this view provides tools for forecasting usage and increasing allowance. This view is usually reserved for leaders within the business, and is what we’ll cover in this article.
Accessing the Dashboard
The Dashboard is accessed via the Main Menu.
What the Dashboard Shows
If you are part of a user group with the permission "View All Agentic Usage", then when viewing the Dashboard, the data will represent all users.
You'll get an overview of what agents have been running.
The tiles a the top of the screen show the totals for This week, this month and all time, including how many active instances their are across the business.
You can expand any of those instances to see the data for just that agent.
What Counts as an AI Call
There are three separate meters to be aware of:
If you are using automation credits via the Sequencer, this is unchanged
It’s 1 per sequence, end to end. If a trigger kicks off a chain of agents automatically, formatting, summarizing, scoring, all off one shortlist move, that entire sequence counts as a single credit, not one per agent inside it.
Then we have Agentic Credits – they are one per agent run, when it's triggered.
This means, if a user screens twenty-two applicants through the Alignment Score Agent, that's twenty-two actions, not one. It adds up faster than people expect on a high-volume day, so it's worth knowing where you stand.
Finally, we have Eva prompts, where it’s one credit per prompt.
Monitoring Agents
Every agency's allowance looks different depending on your plan, how many users you have, and which functionality you're actually using, so rather than memorizing a number, know that your dashboard is always the current, accurate answer.
But it's more than a cost meter. It's a real window into how your team is actually adopting and using this, and it's worth using it that way, not just checking it before a bill comes due.
The Usage Dashboard shows current usage, who's driving it, and a forecast for the rest of the month. Check it regularly rather than waiting for a limit to hit.
That per-user breakdown is the part worth building a habit around.
Find your early adopters, and put them to work.
Look at who's highest on that all-time column. Those are the people who've already figured this out without needing to be told, and they're your fastest path to getting everyone else up to speed. Don't just notice them, use them: ask one of your high-usage consultants to walk the team through how they actually use it day to day in your next team meeting, or pair them informally with someone who's still at zero.
Spot who hasn't started yet, and check in early.
The flip side of that same table: someone showing zero or barely any runs, weeks after this rolled out, isn't a discipline problem, it's usually a "nobody showed me where to start" problem. Catch it early, one short conversation, not a performance note. Ask what's stopping them, half the time it's as simple as not knowing where the button is.
Watch for unusually heavy usage too.
This cuts the other way as well. If someone's running numbers that look far outside the rest of the team, more than a genuinely heavier caseload would explain, that's worth a conversation early, before it becomes a cost surprise at the end of the month. Not an accusation, a quick "hey, walk me through how you're using this," it might be completely legitimate, or it might be an easy fix, like running an agent three times because they weren't sure it worked the first time.
Automation Allowance & Forecast
Pay particular attention to the "Automation Allowance" and Forecast sections of the Dashboard.
The Automation Allowance panel tracks your Automation credits specifically, these are your agent instances that run automatically, and not the manual on-demand runs.
Visible here are the actual number of active automations, monthly allowance and runs, and what's remaining.
The forecast chart projects your month-end total based on what's actually happened so far, so if you're trending toward your limit, you'll see it two weeks out, not on the day it happens.
When you need to upgrade, the dashboard provides the ability to do so directly when you're approaching your allowance, you won't need to go looking for it separately.
Just click the Upgrade Allowance button to be taken to Tools & Settings > Licences & Credits.
You're not locked into upgrading your whole plan tier just because one credit type is running tight, here you can add bundles for each credit type individually, so if you're burning through Automation credits but your Eva Prompt usage is fine, you top up Automation specifically rather than paying for a broader upgrade you don't need across the board.