How AI is Transforming IT Service Desk Automation for MSPs
How We Integrated Google AI Studio, Bomgar, Autotask, and Slack
Recent advancements in AI are reshaping IT service desk automation in ways that are especially valuable for Managed Service Providers (MSPs). Beyond writing assistants like Grammarly, AI now enables automation of tasks that once required hiring developers or making significant investments.
At Network Thinking Solutions, we’ve been experimenting with these tools and building them into our workflows. Having a strong internal team gives us a unique advantage, but what excites me most is that some of the biggest long-standing pain points in IT support are finally solvable.
The Ticketing Pain Point Every MSP Faces
One of the most time-consuming parts of IT support has always been finalizing tickets after remote support sessions. Documenting everything thoroughly is critical, but it pulls technicians (myself included) away from helping the next client. Skipping documentation creates bigger problems later, yet doing it right has always been tedious.
Why We Still Use Bomgar (BeyondTrust) for Remote Support
For over a decade, we’ve relied on Bomgar (well before BeyondTrust acquired them). We initially chose it because it provided features our RMM couldn’t:
Recorded sessions for accountability
Linux support for technicians
Built-in chat for client communication
Safe Mode with Networking for reboots
While our RMM has remote access, Bomgar’s screen recording feature remains invaluable. It helps technicians remember what was done, gives clients visibility, and allows us to quickly review mistakes and retrace steps. After each client session, the details of the session and a link to the recording are put into an Autotask ticket via API. The client then waits for our team to update the recording ticket with our notes and summary of the resolution.
AI and IT Service Desk Automation: The Breakthrough
When OpenAI first launched ChatGPT, my immediate thought was: could AI watch our session recordings and generate ticket summaries automatically?
The challenge: AI tools could summarize transcripts, but our Bomgar recordings didn’t produce transcripts. I tested ChatGPT, Claude, Gemini, Mistral, DeepSeek, and others, with no success.
That changed when I discovered Google AI Studio, which can summarize YouTube videos. Uploading a Bomgar session video produced accurate, concise summaries. From there, the team went further:
Upload Bomgar recordings directly via Google AI Studio’s API (no YouTube needed)
Generate session summaries instantly
Feed those summaries into Autotask tickets
Notify technicians in Slack for quick review
How We Integrated Google AI Studio, Autotask, and Slack
Because we already had Slack and Autotask integrated, connecting everything was straightforward. Now, here’s how the process flows:
Record session in Bomgar (as always).
Send to Google AI Studio API → AI generates a session summary.
Automatically attach the summary to the correct Autotask ticket.
Send Slack notification so the technician can review and finalize.
The result? Remote support ticket automation that saves time, improves accuracy, and keeps both technicians and clients happy. The upload to Google AI Studio is automatically done in the background, and the team member is sent a DM (Direct Message) in Slack when done.
For those wanting to dive deeper, we use Google AI Studio, Gemini (gemini-2.5-pro), Bomgar (Beyond Trust), FFmpeg, SQLite, Slack, Autotask, Python, and GitHub to create the solution we wanted.
Version 2 of this, we plan to merge what we already do today with Slack and Autotask. Once the tech receives the DM, they will be able to adjust Issue Type, Ticket Status, and other key fields in Slack, add any notes, and update the ticket within Slack, without ever going to Autotask to interact with the ticket.
The Results: Faster Ticket Closure and Happier Clients
The impact has been tremendous:
Clients get immediate, detailed updates after remote sessions.
Technicians save time by focusing on support instead of tedious ticket notes.
Documentation is consistent, transparent, and always tied to session recordings.
In short, AI is no longer a “nice-to-have”; it’s becoming essential for MSPs looking to scale IT service desk efficiency.
Here is a screenshot of the completed summary posted to the Autotask ticket. We do a lot of work with Manufactured Housing and the property owners. The example below uses the website manageamerica.com and shows how we help our clients in all facets of support, such as troubleshooting a website and then helping to open a ticket. The details pulled from the recorded sessions are invaluable.
👉 If you’re interested in how we’re applying AI automation for MSP workflows, or want to see what’s possible in your environment, visit Network Thinking Solutions to learn more.
