Q1. Could you start by giving us a brief overview of your professional background, particularly focusing on your expertise in the industry?

I have spent my entire career at the intersection of enterprise technology and the public sector, which gives me a distinct vantage point: I understand how cutting-edge technology is built and positioned, and I understand the specific, friction-filled reality of getting that technology adopted inside government.

I spent nearly a decade at Cisco in public sector sales before moving to Google Cloud, where I served as Partner Development Manager for SLED Central, covering state, local, and education customers across Illinois, Wisconsin, and Minnesota. At Google, I drove over $10M in GCP pipeline through a GenAI outreach campaign, managed $2M+ in partner funding to accelerate Gemini pilots, and worked daily at the convergence of AI product capability, government compliance architecture, and channel economics. I tracked Gemini's evolution from an add-on SKU to a bundled enterprise platform, worked through the practical realities of FedRAMP High authorization, and navigated the genuine complexity of moving agencies from AI curiosity to production commitment.

I am now back at Cisco, which means I sit with a rare dual lens: I have seen the Google Cloud and AI ecosystem from the inside, and I have spent the better part of a decade understanding how Cisco's networking, security, and infrastructure portfolio fits into the same government accounts. That combination, cloud-native AI on one side and legacy infrastructure on the other, is where I spend most of my thinking today.

 

Q2. From a budgetary standpoint among major SLED agencies right now in mid-2026, are we seeing a shift in the budget allocation for modernizing edge networking (routing/switching) and pure-play cloud migration?

The honest answer is that it is not a clean either/or, and agencies that are being told it is a clean either/or are being misled by vendors with an obvious interest in the answer.

What I am observing is a bifurcation. Larger state agencies and university systems with dedicated IT leadership are accelerating cloud migration, particularly for productivity workloads and AI-adjacent use cases. The Workspace and M365 installed base decisions of the last few years effectively pre-committed those agencies to cloud-first architectures for collaboration, and AI capability is now pulling more workloads into that orbit.

But the edge networking budget has not collapsed in response. If anything, the AI workload push is creating new edge requirements, particularly around data movement, low-latency inference at the agency edge, and the network capacity needed to support video-heavy collaboration and AI-assisted workflows at scale. Agencies are discovering that cloud migration does not eliminate the network conversation; it changes it. Routing and switching modernization is increasingly justified not as legacy maintenance but as AI readiness infrastructure.

The agencies getting into trouble are the ones that approved cloud migration budgets without a corresponding refresh of the network layer underneath. That mismatch is one of the more predictable friction points I see in mid-2026 SLED deployments.

 

Q3. As agencies attempt to scale up data-intensive analytics and automated software workflows, how severely are unexpected hybrid-cloud network latency and data movement (egress) fees impacting their deployment timelines?

More severely than most agencies anticipated, and more severely than most vendors communicated upfront. This was a genuine pain point I encountered repeatedly during my Google tenure, and it has not been resolved.

On latency: hybrid cloud architectures where sensitive data must stay on-premises but analytics workloads run in the cloud create round-trip latency that becomes operationally significant at scale. Agencies running large dataset analytics, particularly health and human services or justice agencies with complex case management workflows, hit latency walls that their PoV environments never revealed because the pilot data volumes were too small.

On egress fees: this was the more acute budget problem. Government agencies plan annual budgets with fixed line items. Cloud egress fees are consumption-based and scale with data movement in ways that are genuinely difficult to forecast accurately. I saw agencies whose initial cloud consumption estimates were off by 40 to 60% once production-scale data movement began. That variance creates mid-year budget crises in environments where reprogramming funds is slow and politically difficult.

The practical impact on deployment timelines was real: agencies that hit unexpected egress cost overruns in year one became significantly more conservative about expanding workloads in year two. It was one of the more underappreciated risks in the AI and cloud modernization narrative that vendors were telling in 2024 and 2025.

 

Q4. Based on your experience, what realistic percentage of an initially generated AI sales pipeline actually converts into multi-year, production-scale cloud consumption contracts?

The number that gets quoted internally is almost always more optimistic than reality, and I say that having lived on both sides of that conversation.

From what I observed across my SLED Central territory at Google, a realistic conversion rate from initial AI pipeline engagement to meaningful production commitment within 12 months was in the 30 to 40% range. That means 60 to 70% of deals that entered the pipeline as genuine opportunities, not just marketing-qualified leads, stalled before reaching production scale.

The stall reasons were consistent: budget cycle misalignment, ATO and security review timelines that extended beyond the initial enthusiasm window, identity integration complexity that surfaced during implementation, and change management gaps where the technology was deployed but the workforce adoption never followed. That last one is underappreciated. An agency can sign a multi-year contract and still generate minimal cloud consumption if the organizational change management is not funded and staffed.

For multi-year production-scale consumption contracts specifically, I would put a realistic conversion at 20 to 30% of initial pipeline within an 18-month window. The agencies that converted reliably shared a common characteristic: a strong internal champion at the program level, not just IT, who had budget authority and organizational motivation to drive adoption past the pilot stage.

 

Q5. From your experience bridging cloud-native partner ecosystems with legacy government requirements, how much influence do regional systems integrators and VARs actually hold today?

More than most cloud vendors publicly acknowledge, and the smart vendors know it.

The narrative from hyperscaler sales organizations is often that the cloud platform is the product and the partner is the delivery mechanism. That framing fundamentally underestimates what regional SIs and VARs actually do in government accounts. They hold the procurement relationships that predate the current technology cycle by years. A CDW rep covering mid-size Illinois counties has relationships with procurement officers and IT directors that Google or Cisco field reps cannot replicate with a quarterly visit. That relationship equity is real, and it transfers to technology decisions.

In SLED specifically, the VAR's influence over the final vendor selection was often more determinative than the vendor's own field motion on accounts below the named account threshold, which means the majority of accounts. Partners also controlled the implementation timeline, the training quality, and the change management depth that determined whether a deployment became a reference story or a cautionary tale.

Where I see VAR influence eroding is on pure commodity resale, which is exactly where cloud marketplaces are winning. But on complex government deployments involving compliance configuration, identity integration, and workforce adoption, the regional SI's influence is not declining. It is, if anything, becoming more important as the technical complexity of AI deployments increases.

 

Q6. As cloud marketplaces aggressively consolidate enterprise software procurement, how are regional Midwest VARs adjusting their service-delivery business models to preserve margin, and what is the impact of this shift?

The partners who saw this coming adjusted early and are doing well. The partners who treated marketplace consolidation as a future threat rather than a present reality are under genuine margin pressure.

The adjustment pattern I observed, most clearly with CDW and their Amplified for Education and public sector practices, was a deliberate migration away from license resale margin toward annuity services revenue. The resale margin on a Google Workspace or Microsoft M365 seat was never the business. The business was the implementation engagement, the managed service contract, the professional development program, and the ongoing advisory relationship that the resale relationship made possible.

The partners who are thriving in mid-2026 have built repeatable service offerings around AI adoption, specifically: change management programs, AI governance frameworks, prompt engineering training for agency staff, and managed AI operations that monitor usage, governance drift, and compliance posture on behalf of agencies that do not have the internal capacity to do it themselves. That is a service that a cloud marketplace cannot commoditize.

The partners who are struggling are the ones whose revenue was disproportionately dependent on volume resale and who did not invest in building service delivery IP when the margin was still good enough to fund that investment. In the Midwest specifically, smaller regional VARs without a differentiated services practice are facing a difficult next 24 months as marketplace procurement continues to disintermediate the transaction.

 

Q7. If you were an investor looking at companies within this space, what critical question would you pose to their senior management?

The question I would ask is: what is your net revenue retention rate specifically among customers who completed a production deployment, and how does that compare to customers who stalled at pilot?

The reason I would ask that question is that it cuts directly through the AI hype cycle in a way that pipeline numbers and logo counts do not. Every vendor in this space can show impressive pipeline growth and a growing list of pilot customers. What that metric does not reveal is whether the underlying product and deployment model actually generate durable consumption growth once an agency is running at scale.

In my experience, the gap between pilot and production is where the real economics of the business are determined. A company with strong net revenue retention among production customers has solved the hard problems: the compliance configuration is repeatable, the change management model works, the partner ecosystem is generating genuine adoption, and the product is delivering enough measurable value that agencies expand rather than contract.

A company with weak net revenue retention among production customers, regardless of how impressive the new logo announcements look, has a fundamental problem that more sales investment will not fix. In the SLED AI market specifically, where budget cycles are slow and switching costs are real, a stalled production customer is often a lost customer at the next renewal. I would want to understand that dynamic before making any investment decision in this space.