Escaping the IT “Doom Loop” in MSPs – with Leaders from Xurrent, Impact Networking, and Savaco

This interview analysis is sponsored by Xurrent and was written, edited, and published in alignment with our Emerj sponsored content guidelines. Learn more about our thought leadership and content creation services on our Emerj Media Services page.

Enterprise IT and service operations teams face structural pressures that are remarkably consistent across industries. Despite advances in automation and observability, leaders continue to report chronic firefighting, repeated incidents, and operational drag that restricts modernization. 

Recent research illustrates the scale of the challenge. According to a 2025 World Economic Forum survey, 54% of large organizations identify supply chain challenges as their most significant barrier to achieving cyber resilience, driven by the complexity of modern, interconnected systems. 

As infrastructure grows ever more distributed, so do the dependencies, hand-offs, and failure points that service teams must manage — often without the automation or visibility necessary to meet demand.

In fact, Harvard Business Impact’s 2025 Global Leadership Development Study underscores the dilemma, noting a greater perceived importance of leaders’ ability to function in a constantly changing environment, up from 58% in 2024 to 71% in 2025.

To explore these challenges, Emerj’s ‘AI in Business’ podcast recently hosted a series of episodes to break down how Managed Service Providers (MSPs) are thinking about Enterprise Service Management (ESM), automation, and the widespread integration of AI into these systems. 

Sponsored by Xurrent and hosted by Emerj Editorial Director Matthew DeMello, Phil Christianson, Chief Product Officer at Xurrent; Steve Taczala, VP of Service Operations at Impact Networking; and Dirk Michiels, CEO of Savaco, each shares their viewpoints and offers guiding frameworks for enterprise leaders grappling with their own AI ecosystems.

Drawing insights from their experiences in MSP operation, this article examines how AI — when grounded in disciplined workflows, governance, and enterprise-ready practices — can shift service operations from reactive firefighting to proactive resilience, with particular focus on:

• Ending the recurring “Doom Loop” in incident response: Using AI-supported triage, routing, and auto-generated post-incident timelines to shorten recovery cycles, reduce repeat issues, and create a more predictable, resilient incident-response process.
• Reducing ticket noise in MSP and enterprise environments: Applying AI-driven noise suppression, classification, and prioritization to eliminate low-value alerts and improve signal quality.
• Simplifying infrastructural complexity in mid-market IT: Standardizing and streamlining core workflows, strengthening governance and alignment, and targeting AI deployment at high-friction bottlenecks to deliver robust business value.

Ending the Recurring “Doom Loop” in Incident Response

Episode: Breaking the Doom Loop in IT Service Management – with Phil Christianson of Xurrent

Guest: Phil Christianson, Chief Product Officer, Xurrent

Expertise: Product Management, ITSM Platforms, e-Commerce Technology, AI-Driven Software Solutions

Brief Recognition: Phil drives product strategy and innovation as Chief Product Officer of Xurrent’s AI-powered IT service management platform. Earlier at Wayfair, he oversaw dynamic pricing for 25 million products, following over a decade of building enterprise software solutions across a variety of enterprises.

Christianson describes a dynamic that persists even in technologically mature environments: major incidents cycle through the same patterns, with the same teams pulled into urgent war rooms, often without long-term resolution.

As he explains, IT often absorbs disproportionate responsibility: “Regardless of where the problem came from, IT teams often sit in the middle of those war rooms.”

The deeper issue is what Christianson and IT professionals across sectors refer to as the “doom loop” — a cycle in which organizations excel at rapid incident response but struggle to translate insights into durable improvements. Post-mortems are documented but not always operationalized, and chronic issues reappear in slightly altered forms. 

“We think of the doom loop as focusing [too heavily] on the war room: you focus on getting people in, fixing the problem, and getting out. But now that we have these tools, and we’ve made the job of the commander faster, easier, and better, we believe that the next phase is longer term resiliency; how do you take what happened in the war room, make sure that the tasks coming out of the post-mortem are sent to the teams, and that they’re held accountable to those?”

– Phil Christianson, Chief Product Officer at Xurrent

The loop erodes resilience, but with the AI tools now available, mechanisms to drive follow-through allow teams to move beyond short-term reaction, capturing the structural gains needed to reduce incident volume.

Christianson emphasizes that while war rooms will never disappear entirely, AI can fundamentally change how they operate. Traditionally, commanders sift through Slack threads, alert logs, and phone updates to reconstruct the incident timeline – a process, Christianson notes, that can be easily automated with AI.

As Christianson explains: “You can have things like post mortems auto-generated from a view of the entire conversation, it can look across the conversations, the alerts that came in – and it can construct a timeline.”

Critically, these changes reduce human error and compress documentation tasks to seconds. Christianson highlights the broader value: “AI can be used to make the war room better, faster, more efficient, and allow you as a company to think beyond the war room – because you now have the time.”

The result is not only faster remediation but also improved long-term resilience, as tasks emerging from post-mortems are captured, routed, and tracked rather than lost in ad hoc documents.

Reducing Ticket Noise in MSP and Enterprise Environments

Episode: Fixing Ticket Noise with AI in Enterprise MSP Operations – with Steve Taczala of Impact Networking

Guest: Steve Taczala, VP of Service Operations, Impact Networking

Expertise: IT Service Operations, MSP Management, End-User Support Services, Workflow Optimization, Process Improvement

Brief Recognition: Steve oversees service delivery for Impact Networking’s nationwide managed service provider operations. Bringing nearly two decades of IT operations leadership across firms such as SunGard and Synoptek, Steve has a track record of building customer-focused global support organizations and streamlining IT workflows for greater efficiency.

In managed service provider environments, recurring noise interferes with both accuracy and speed. Taczala emphasizes the magnitude of the problem: “One of the main issues around service desks in MSPs today is ticket noise. They are the first gateway to all requests.”

Misconfigured alert thresholds, redundant communications, and non-actionable acknowledgments, such as users replying “thank you” to a closed ticket, all end up in the queue. As Taczala explains, “That ticket’s not actionable… but it’s ingested just like any other ticket.”

In large MSP or enterprise environments, noise can account for 10–15% of total ticket volume. Taczala observes, “As you get more into the larger MSP enterprise space, 10 to 15% of noise could be north of 1,000 plus tickets, and that causes far too much distraction.”

That amount of volume dilutes teams’ ability to detect accurate signals, slows time-to-response, and increases the likelihood that low-impact issues crowd out severe incidents. AI-enabled classification, however, helps organizations suppress non-actionable tickets before they hit the queue.

AI models can detect patterns indicating non-actionable requests — such as closed-ticket replies or alerts that appear without meaningful thresholds — and automatically suppress or reroute them. Taczala illustrates one typical example: “You close a ticket… the client replies with a thank you… and it would ingest another ticket. But that ticket’s not actionable.”

Automating such early filtering creates two immediate benefits:

• Reduced workload for Level 1 and service desk teams
• Improved visibility into real issues that require human intervention

Noise reduction is not only an efficiency gain; it improves the signal-to-noise ratio that underpins reliable incident detection across the business.

Once noise is suppressed, AI can help organizations move beyond manual triage and achieve more consistent routing. Today, many service environments still operate on first-in-first-out habits or inconsistent categorization practices that dilute response time for high-impact issues:

“A platform that manages first-in-first-out tickets doesn’t have the means to prioritize. So all tickets are created equally. That causes confusion. It causes misprioritization of client tickets. You would be working on a service request before you actually work on a downed device that’s causing an outage at one of your locations.

So before you migrate, you need to have a basic understanding of what your platform wants, what you need that platform to do, to develop the best customer experience.”

– Steve Taczala, VP of Service Operations at Impact Networking

AI systems can:

• Interpret device criticality based on metadata
• Identify incident type and recommended assignment
• Prioritize outages versus routine service requests
• Route based on skill level, not just availability

As Taczala explains, recovery time often hinges more on routing speed than the complexity of the fix: “A fix could take 15 minutes, but if it takes eight hours to get to that person, you have an eight-hour and 15-minute outage.”

AI-based routing ensures that tickets land with the right individual or team the first time, closing the gap between detection and remediation.

Simplifying Infrastructure Complexity in Mid-Market IT

Episode: Overcoming Cloud Complexity in Mid Market Operations – with Dirk Michiels of Savaco

Guest: Dirk Michiels, CEO, Savaco

Expertise: IT Service Management (ITSM), Digital Transformation, Data Analytics, AI Solutions, Managed Services

Brief Recognition: Dirk leads Savaco’s digital transformation and managed services business as CEO. Previously, he served as CEO of Ferranti Computer Systems and later as CEO of the AI startup Tangent Works. Dirk brings decades of enterprise IT leadership experience, including over a decade as a Savaco board advisor.

Mid-market organizations face a different but equally significant challenge: managing enterprise-level complexity without the resources to match. Michiels summarizes the situation succinctly: “There is a consistent and persistent misalignment between IT and business, even though people have put efforts in modernization.”

Hybrid and multi-cloud architectures, legacy systems, inconsistent integration patterns, and tool fragmentation create operational friction that slows modernization. Michiels notes that traditional change approval processes are “just not fit for the task” in environments where speed and reliability are essential.

The result is a widening gap between business expectations — rapid deployment, high service reliability — and what IT teams can deliver with existing systems and processes.

While IT service operations are the natural starting point for AI-enabled transformation, the same workflows, principles, and capabilities extend across the broader enterprise. Leaders in every function — HR, Finance, Facilities, Procurement — depend on ticketing, triage, approvals, and service delivery. 

In organizations where these processes remain fragmented across tools and teams, service experiences become inconsistent and operational visibility degrades. The misalignment Dirk describes between IT and business is magnified outside the IT space, where:

• Workflows often depend on separate systems
• Data is siloed across functions
• Approvals are manual and inconsistent
• Service quality varies based on department maturity

These conditions make it difficult for businesses to operate with the predictability and responsiveness senior leadership expects.

ESM, however, provides a unified operational model. As Michiels notes, comprehensive AI tools like Xurrent support integrated workflows across multiple service functions, allowing organizations to manage hybrid infrastructure, coordinate change, and automate routine requests from a single platform. 

He explains that Savaco has adopted the model he describes across its MSP practice, integrating Xurrent’s platform with monitoring tools such as LogicMonitor and Microsoft services to support end-to-end operational workflows.

The same architecture that supports incident, change, and problem management in IT can support:

• HR onboarding
• Finance approvals
• Vendor and facilities requests
• Cross-department service catalog workflows

These elements create a coherent, shared operational fabric that reduces friction, increases transparency, and improves employee experience across the enterprise.

Michiels also notes the importance of alignment and governance to derive tangible value from AI, telling the Emerj executive podcast audience that “if you don’t align transformation initiatives with process improvements and goals, then AI will not be the business value driver, it will just be a technology experiment.”

For AI to deliver sustained value in service operations or ESM, organizations need: 

• Clear policy boundaries
• Defined approval processes
• Transparent data handling
• Continuous executive sponsorship
• Guardrails for cost controls and responsible use

Without these, AI tools may accelerate operational risk rather than mitigate it.