In Part One of this series, we talked about the canonical data layer: why enterprises struggle to get a consistent picture of their software IT asset landscape even when the data technically exists, and how a shared metadata layer is the prerequisite for anything useful downstream.
This part is about what happens next.
Once you have a clean, normalized, reconciled view of your licenses, users, usage, and spend, the question is: what do you do with it? The answer most platforms give is a dashboard. Charts, filters, export to Excel. Maybe an alert that fires when a threshold is crossed. That is table stakes. It is not the same as intelligence.
The gap between having data and making decisions from it is still enormous in most ITAM organizations. An analyst can pull a utilization report for Salesforce and see that 200 seats are active out of 600 licensed. That is a useful data point. But it does not tell them which 400 seats to harvest, which of those have dependencies, what the contract terms allow, when the renewal is, what the financial impact is, or what to do first. Getting from the data point to the actual decision still takes days of work.
That is the gap intelligent agents are built to close.
Meet Your New Team of Autonomous Specialists
The word "agent" gets used loosely. In the context of Asato, an agent isa specialized IT asset component that does a specific job autonomously: ingesting data from a source (e.g contract pdfs), running a reconciliation, monitoring for a condition, or executing an approved action against an external system. Agents are not general-purpose chatbots. They are narrow, purpose-built, and they run continuously in the background. Think of them as a team of digital specialists working for you 24/7.
- The Entitlement Agent – Watches license entitlements and flags anomalies against contract terms.
- The Identity Agent – Monitors identity provider activity to track who last logged into which application.
- The Procurement Agent – Cross-references procurement records with vendor portals to catch seat count mismatches before they become audit findings.
None of them wait to be asked. Agents fix the ultimate failure mode of traditional ITAM: timing. Most teams discover waste only at renewal time or during an audit. Agents keep analysis live, giving you answers when you actually have time to act.
Four Ways Asato Categorizes Insights
Not all ITAM findings are the same kind of problem. Asato organizes these agent-driven insights into four categories, and the distinction matters because each one calls for a different response.
| Insight Type |
What the agent is looking for |
Example Finding |
| Utilization |
Utilization Tracks who is using what, how often, and from where. Seat counts versus active logins versus feature engagement. 142 Microsoft 365 E5 accounts have not logged in for more than 90 days. At $57 per seat per month, that is $8,094 in recurring waste. |
Utilization Tracks who is using what, how often, and from where. Seat counts versus active logins versus feature engagement. 142 Microsoft 365 E5 accounts have not logged in for more than 90 days. At $57 per seat per month, that is $8,094 in recurring waste. |
| Shadow IT |
Shadow IT Catches unapproved apps appearing in web proxy or firewall logs before they become security leaks. Web proxy logs show 22 users in the Sales division regularly accessing Monday.com. There is no corporate contract, no SSO provisioning. |
Shadow IT Catches unapproved apps appearing in web proxy or firewall logs before they become security leaks. Web proxy logs show 22 users in the Sales division regularly accessing Monday.com. There is no corporate contract, no SSO provisioning. |
| Rationalization |
Rationalization Identifies portfolio overlaps. (which tools serve the same purpose for the same users, based on usage data). A business unit is paying separately for Webex for 300 users while the company already has an enterprise Zoom license with 340 inactive seats. Usage data shows 240 of those Webex users already have Zoom access that they are not using. |
Rationalization Identifies portfolio overlaps. (which tools serve the same purpose for the same users, based on usage data). A business unit is paying separately for Webex for 300 users while the company already has an enterprise Zoom license with 340 inactive seats. Usage data shows 240 of those Webex users already have Zoom access that they are not using. |
| Optimization |
Optimization Restructures existing contracts for tier downgrades, renewal timing, and seat reallocation. 200 Salesforce Sales Cloud seats in the Customer Success division are used only to view account records and log calls. Downgrading them to Sales Cloud Professional saves $72,000 at renewal without removing any functionality those users need. |
Optimization Restructures existing contracts for tier downgrades, renewal timing, and seat reallocation. 200 Salesforce Sales Cloud seats in the Customer Success division are used only to view account records and log calls. Downgrading them to Sales Cloud Professional saves $72,000 at renewal without removing any functionality those users need. |
Most ITAM tools blend these insights together, which makes prioritization harder than it needs to be. Keeping them separate means the right person gets the right finding and knows exactly what kind of decision is being asked of them.
Going From Insight to Action: The Four Steps
Insight without a path to action is just a report with a nicer interface. The part that actually changes the business is what happens after the finding. Asato bridges this gap using a structured four-part action framework as illustrated here.
| Action Type |
What Happens |
Example |
| Notify |
Notify Surfaces a finding to the right owner with full, actionable context. The ServiceNow app owner receives a summary showing 60 unused ITSM Pro licenses expiring in 47 days, with a recommended action and the data behind it. |
Notify Surfaces a finding to the right owner with full, actionable context. The ServiceNow app owner receives a summary showing 60 unused ITSM Pro licenses expiring in 47 days, with a recommended action and the data behind it. |
| Recommend |
Recommend Presents a ranked, justified proposal for review. Asato surfaces a license harvest plan for Adobe Creative Cloud: 38 seats flagged as inactive, ordered by inactivity duration, with downstream dependency checks already run. |
Recommend Presents a ranked, justified proposal for review. Asato surfaces a license harvest plan for Adobe Creative Cloud: 38 seats flagged as inactive, ordered by inactivity duration, with downstream dependency checks already run. |
| Automate |
Automate Once approved, an agent executes the workflow directly against the relevant system, eliminating manual steps. After an ITAM lead approves the Adobe harvest, Asato deactivates the 38 licenses in the vendor portal and opens a confirmation ticket in Jira. No manual steps. |
Automate Once approved, an agent executes the workflow directly against the relevant system, eliminating manual steps. After an ITAM lead approves the Adobe harvest, Asato deactivates the 38 licenses in the vendor portal and opens a confirmation ticket in Jira. No manual steps. |
| Govern |
Govern The agent applies ongoing policies and enforces a defined state continuously. Any newly detected orphaned account, meaning an account still active after an employee offboarding, is flagged within 24 hours and queued for deprovisioning review. |
Govern The agent applies ongoing policies and enforces a defined state continuously. Any newly detected orphaned account, meaning an account still active after an employee offboarding, is flagged within 24 hours and queued for deprovisioning review. |
The Golden Rule: Humans Stay in the Lead, not just in the loop. The critical design principle here is that automation never runs without approval. Every Automate action sits behind a human review step. The agent does the legwork, presents the proposal with supporting evidence, and waits for a decision. When the decision is made, execution is fast and logged. But the human stays in the lead. The value of automation here is not removing the human, it is removing the manual work that precedes the decision so the human can focus on the judgment call rather than the data gathering.
This is the right architecture for IT asset decisions, where the consequences of a mistake (accidentally revoking access for an active user, or triggering a contract penalty by harvesting the wrong seat type) are real.
The Bottom Line for Enterprise Leaders
This shift is not a technology story; It is an organizational upgrade designed for your ITAM teams.
Right now, most enterprise ITAM teams are operationally overwhelmed. They understand the problems clearly, but legacy tools leave them with little time for strategic thinking. Continuous agent monitoring, structured insight categories, a clear action framework, and a natural language interface do not make the ITAM function smaller. They make it exponentially more capable of doing the real work.
Teams that move in the direction of adopting a continuous, agent-driven framework stand to gain immense first-mover advantage. Not because the technology is magic, but because they will be making decisions from current, complete data while their peers are still assembling last quarter's picture in a spreadsheet.
That gap compounds over time. Renewals negotiated with better data produce better contracts. Rationalizations done with full portfolio visibility produce cleaner, cheaper toolsets. Access governance run continuously produces fewer audit findings and fewer surprises. The value is not in any single decision. It is in the steady accumulation of better decisions made faster, by a team that finally has the right foundation to work from.
.png)
