Campai achieved AI assistance across nearly 70% of service desk tickets handled through Thread.

Campai is a 50-person managed services provider based in Zoetermeer, Netherlands, serving small and medium-sized businesses across the country. Most of their customers have between 25 and 150 employees, spanning multiple industries with a notable concentration in the Dutch maritime and boating sector. Campai handles full-stack IT management (workstations, Microsoft environments, security, modern workplace) for clients ranging from three to three hundred workstations.

Before Campai got serious about Thread, their service desk ran the way most MSPs do: email, phone, ad-hoc tickets, and a lot of repetitive back-and-forth. When they brought Thread in, they saw a bigger opportunity than just faster ticket handling. They wanted an AI layer that sounded like Campai, operated in Dutch, and could automate the multi-system workflows their techs were doing by hand every week.

The Core Problem: Every predictable request, including routine ones like "I'm going on holiday to Spain next week, please unblock my account," consumed a technician's time from first contact through resolution, with no meaningful leverage from automation. And Campai wasn't willing to solve that with an off-the-shelf bot voice their customers would reject.

"We programmed it all like a human. The tone of voice, everything. People know they're talking to a bot, but they don't want to know."
— Janneke Post, Modern Work Consultant, Campai

Alongside voice, the team had specific operational pain:

• Predictable, multi-step workflows like travel geo-block approvals were being handled manually every single time, across multiple systems: Autotask, Microsoft Entra ID, and Rewst.
• Customers often forgot to request access changes before they traveled, which meant support tickets from locked-out users in airports and hotels.
• Technicians were spending cycles on requests that followed the same pattern every time, with no way to templatize the response without losing the human feel customers expected.

Campai wanted automation and a human customer experience. Thread gave them the tooling to build both.

Thread was deployed as Campai's conversational AI service desk, connecting Microsoft Teams to Autotask and giving the team a programmable intent layer to automate multi-system workflows that would otherwise require custom middleware.

Campai's adoption played out in two phases.

First, they made the agent sound like Campai. Because Thread exposes the triage agent's prompts and tone as something partners can shape themselves, the Campai team authored their own voice and built in the clarifying steps their techs would normally ask. The agent checks the computer number if someone's complaining about their machine, runs a first-pass analysis, and then offers the user a choice: "Do you want to troubleshoot with me or do you want an engineer?"

"We wanted it to feel like a nice service desk experience. Friendly. So the agent does a lot of checks for us, analyzes the issue first, and then asks: do you want to troubleshoot with me, or do you want an engineer?"
— Janneke Post, Modern Work Consultant, Campai

 

Second, Campai dug into Thread's intents: structured prompts that let a single team member build agent-driven workflows without waiting on product engineering. Janneke, who joined Campai as a Technical Account Manager and moved into a Modern Work Consultant role in September 2025, started shipping intents during a two-week stretch when most of the team was on holiday. One of those intents became the proof point.

Today Thread sits in front of Campai's Autotask service desk as the intake, triage, and automation layer. Every chat ticket comes in through Microsoft Teams. The triage agent handles first response, gathers context, and (for a growing set of request types) executes the full workflow before a technician ever sees it.

Day in the Life: The Holiday Geo-Block Flow

A Campai end user opens Teams and sends a message: they're going to Spain next week. Thread's intent picks it up, asks where they're going and on which dates, and replies in the language of the destination country ("¡Hola!"). It routes an approval request to the user's manager via Teams or email. Once approved, Thread hands off to Rewst, which adds the user to the correct Microsoft geo-exclusion group so they can sign in from Spain. Thread creates the ticket in Autotask, confirms to the user that they're unblocked until their return date, and schedules a closing ticket for the day they come back, which automatically removes them from the exclusion group and checks in to make sure nothing went wrong.

Since August 2025, this one intent has run 90 times end-to-end. Zero technician minutes per request.

"For the user, it's maybe two minutes in Teams. For the approver, it's one click. Nobody's sending emails asking us to unblock their account from a hotel in Italy anymore."
— Janneke Post, Modern Work Consultant, Campai

The team also has a standing working relationship with Thread's product group, which lets them shape the platform around Campai's actual workflows.

"Sometimes we have strange requests. 'Is this even possible?' And the Thread team just turns it into a feature request. I never feel like I'm bothering them. I think they actually like that we're this committed."
— Janneke Post, Modern Work Consultant, Campai

Campai Processed Nearly 36,000 Tickets Through Thread While Rolling Out Across Their Client Base

"Since August, this one flow has run ninety times. That's ninety tickets that used to take a technician's time. And now they don't."
— Janneke Post, Modern Work Consultant, Campai

In 2024, Campai handled 922 tickets through Thread, almost none of them AI-assisted. Average resolution time sat around 11 days.

Through 2025 and into 2026, those numbers look completely different:

• 35,969 total tickets handled through Thread
• 16,052 AI-assisted tickets: a 45% AI coverage rate
• 1.2 minute average first AI response
• 4,365 minutes average AI-led resolution time, versus 5,031 minutes for manually-handled tickets, a measurable efficiency gain even within the post-Thread period
• 3,297 automated triage sessions
• 56 average sentiment score, an indicator that Campai's customers are responding well to the conversational flow the team designed

The holiday geo-block intent is the most visible example, but it's one of several automations the team has shipped on top of Thread.

Did Thread help Campai scale without hiring? Yes. Campai processed nearly 36,000 tickets through Thread in 2025–2026, compared to 922 in 2024 as they rolled out to more of their client base, all with the same core team.

Did Thread reduce manual work for technicians? Yes. 45% of tickets now receive AI-assisted handling, and specific workflows like travel approvals are fully automated end-to-end.

Did Thread help Campai serve customers in multiple languages and regions? Yes. Thread's intents let the team respond to EU travel scenarios in the destination country's language, and Campai is now part of Thread's Voice International Cohort helping shape GDPR-compliant voice support for European MSPs.

Campai's story isn't about replacing technicians with AI. It's about an MSP that stopped treating their AI service desk like a chatbot and started treating it like a programmable layer between their customers and their PSA. One person, given time and the right tool, built a cross-system workflow that now runs itself ninety times a year and counting. The service desk got faster, more consistent, and more human-feeling in the process, not less. The Automation Leader of the Year award was a recognition of that work, not a starting point.

Takeaway: Campai shows that MSPs on Autotask can scale ticket volume, automate multi-system workflows, and preserve a human customer experience by deploying Thread as their AI service desk layer, without replacing their existing tools or retraining their team.