Challenge
Tickets on Interware’s service board were claimed on a first come, first served basis. Every technician monitored the board and grabbed what they saw, so triage consumed real capacity, and assignment logic was inconsistent. On the monitoring side, four Tier 1 technicians manually reviewed hundreds of alerts per day, including flapping devices and false positives.
Solution
Interware deployed ConnectWise AI Agents™ within its existing ConnectWise PSA™ environment. The team started with automated triage and dispatch on the service board and with alert analysis on the monitoring side. They also use ConnectWise AI Assistant™ to provide nightly and weekly reporting summaries for service leadership.
Results
Tickets are now assigned automatically upon arrival, without technicians watching the board. Agents bundle or dismiss a large share of alert noise, so technicians focus on the alerts that matter. Leadership estimates that roughly half of a technician’s time has been reclaimed and redeployed to other work.
Interware has run its business on ConnectWise for nearly 17 years. That history left the company with a deep well of ticket data and institutional knowledge. It also gave Interware a clear view of where AI could help. When ConnectWise integrated agentic AI into the ConnectWise Platform™, Interware moved early to apply it to repetitive parts of the workday.
Agentic AI can take on repetitive tasks and do them more efficiently and accurately. For us, the timing was right to invest, so our people can spend more time on higher-value work.
From first come, first served to automated assignment
Interware’s service desk has 10 technicians across three tiers: four at Tier 1, four at Tier 2, and two at Tier 3. The team supports 2,500 endpoints. Before CW AI Agents™, dispatch relied on an open board. Every technician kept it up, and whoever noticed a ticket first assigned it to themselves. The process worked, but it strained attention across the team and left assignment logic inconsistent.
Today, the triage and dispatch agent classifies each incoming ticket and routes it to the technician who already owns the account. More than 1,500 of Interware’s monthly tickets land on the service board where the agent works. The board no longer needs a human to watch it constantly.
Before, dispatch was first come, first served. Everybody watched the board and grabbed tickets as they came in. Now the agent assigns them, and the team does not have to sit there monitoring the queue.
An estimated half a role returned to higher-value work
The savings are clear: four people who used to spend part of every day reading, sorting, and assigning tickets now do none of that manual triage. This additional capacity was redirected to more meaningful work.
Four help desk technicians used to look at tickets and assign them. Now that work is handled automatically. If you add it up across the team, I would estimate we have saved about 50% of a person’s time.
Cutting the alert noise before it reaches a technician
The second major change is on the monitoring side. Interware fields hundreds of alerts a day, and a large share are not actionable. Devices drop and recover within minutes. Alerts duplicate one another. Some are outright false positives. Four Tier 1 technicians used to review every alert. Now the agent first analyzes the stream, bundling related alerts, flagging false positives, and separating what needs attention from what can wait.
Fung estimates that the CW AI Agent’s filter screens out somewhere between 30% and 50% of incoming alerts before a human is involved. The team can then focus on the remaining alerts that need human attention.
A device can go down and come back up two or three minutes later. At that point, the alert is noise. The agents help sort out a lot of those cases, so the team spends time on the alerts that actually matter.
Reporting service leaders can act on
Beyond the queue itself, CW AI Agents now send a nightly digest and a weekly insight summary to service leadership. The reports cover SLA status, activity across the customer base, and volume patterns by customer. That visibility previously required someone to build the report manually.
The nightly digest gives us SLA visibility, and the weekly summary shows patterns by customer. If one customer has a lot of the same kind of ticket, we can ask whether there is something deeper to investigate.
Why AI inside the platform mattered
Interware evaluated third-party AI vendors before committing. The deciding factor was where the data already lived. Interware had almost two decades of tickets and resolutions in CW PSA™ and was leveraging ConnectWise RMM™ on the CW Platform™. Fung also met the product and leadership teams face-to-face at the ConnectWise AI Summit event, which strengthened his confidence in the roadmap.
We have almost 17 years of ticket history and knowledge inside ConnectWise PSA. We looked at other AI vendors, but the ConnectWise team showed real commitment to building this inside the platform. That gave me confidence to invest with ConnectWise and help build it together.
Looking ahead: capacity for higher-value work
Fung sees AI as a natural step for any partner, with a key question: Can AI free up time for people to focus on higher-value tasks? Today, CW AI Agents are helping address triage and alert noise. He expects rapid access to the correct resolution next.
AI is an inevitable step. You cannot keep doing things the old way and pretend it is not happening. Very soon, knowing how to use an agent to help your work will feel as basic as knowing how to use Excel.
Key takeaways
- Start with the work your team repeats every day: CW AI Agents are most valuable when they reduce the routine triage, dispatch, and alert review work technicians already spend time on.
- Use automation to protect technician focus: When AI handles the first pass on tickets and alerts, technicians can spend more time on client-facing service and higher-value problem-solving.
- Make intake more consistent: Standardized classification and routing help tickets move through the service desk with less manual checking and fewer judgment calls.
- Reduce alert noise before it reaches the team: Filtering low-value or false-positive alerts helps prevent unnecessary interruptions and keeps attention on issues that need action.
- Measure capacity, not just activity: The strongest value story is that AI is being used and enabling technician time to be reclaimed for better service work.
- Apply AI to rule-based workflows first: Processes with repeatable patterns, clear criteria, and frequent volume are often the best starting point for agentic automation.
- Build trust through visible workflow relief: Adoption is easier when technicians can see AI removing friction from the work they already know.
