By Guest: Danielle Cook (LinkedIn)
Company: Akamai
Show: 2026 Predictions
Topics: Cloud Native, Kubernetes
The infrastructure decisions enterprises make today will determine whether their AI-powered products feel fast, intelligent, and responsive — or frustratingly slow. That’s the core argument Danielle Cook, Senior Manager at Akamai and CNCF Ambassador, brings to TFiR’s Annual 2026 predictions series. Her outlook is direct: the bar for user experience has been permanently raised, and the organizations that don’t rethink their architecture will fall behind.
📹 Going on record for 2026? We're recording the TFiR Prediction Series through mid-February. If you have a bold take on where AI Infrastructure, Cloud Native, or Enterprise IT is heading—we want to hear it. [Reserve your slot
Akamai’s role in this moment is clear, Cook explains. “AI is fundamentally resetting user expectations. Once people have different experiences — getting responses instantly and wanting real-time information — that raises the bar. Akamai in 2026 is here to help organizations meet and exceed that bar.”
Experience Quality Is an Infrastructure Problem
Cook’s first prediction cuts to the heart of something the industry has long treated as a back-end concern: “Speed and responsiveness are not just performance metrics anymore. They’re defining what good feels like from a product point of view.” In 2026, she argues, infrastructure decisions are product decisions. Organizations that don’t internalize that shift will find themselves losing on experience even when their AI models are excellent.
Her second prediction follows directly: where AI inference runs is becoming a primary design choice. The old model — compute concentrated in a handful of regions or a single centralized data center — is giving way to distributed execution that runs closer to users and data by default. This isn’t a future state. It’s the direction the industry is moving now, driven by the latency demands of real-time AI interaction.
Kubernetes Finds Its Defining Workload
The third prediction is the one Cook finds most significant: Kubernetes is achieving product-market fit through AI. Supported by the recent CNCF annual survey, she points to a clear pattern — AI inference is becoming Kubernetes’ defining workload. “We’re settling into this role for Kubernetes as the standard portable runtime for serving models anywhere we need that distribution,” she says. For an ecosystem that has sometimes been criticized for complexity without clear payoff, this is a meaningful inflection point.
Three Challenges That Will Slow Teams Down
Cook doesn’t stop at predictions. She maps out the challenges that will trip organizations up in 2026. The first is centralized architecture — systems built around central data centers will struggle to support real-time AI interactions at scale. “We’re going to see what we saw with the world wide web. Distance and round trips are going to show up immediately in the customer experience.”
The second challenge is operational. “AI collapses the entire day-two stack into one problem. Security, observability, different pipelines, different GPUs — all of these things come together, and teams are going to have to solve that in one place. They can’t just address each issue singularly.”
The third is a fundamental shift in traffic patterns. Agentic systems don’t behave like traditional request-response architectures. They create continuous, machine-driven interaction loops — and the infrastructure built for the old model won’t hold.
Where the Opportunities Are
The flip side of those challenges is significant opportunity. Cook sees distributed cloud becoming the default application architecture, enabling organizations to run inference closer to users while improving responsiveness and resiliency simultaneously. Real-time personalization — already reshaping retail and travel — extends from this same foundation. Decisions happen in the moment, not after the fact.
At the platform layer, she sees a major opportunity in making Kubernetes effectively invisible. “It’s an opinionated platform that removes the operational burden so your teams can just be deploying AI anywhere.” The cloud native ecosystem — IDPs, golden paths, and community tooling — makes this more achievable than ever.
What Enterprise Leaders Should Do Now
Her advice to enterprise leaders is to design for distribution from day one. “Don’t think about it after the fact. Ask: do I need a distributed cloud? How close to my users do I need to be? Do that at the very beginning of whatever new application you’re building.”
She also urges teams to identify which AI workloads are latency-sensitive early, rather than retrofitting decisions later. And unsurprisingly for a CNCF Ambassador, she advocates standardizing on Kubernetes — not as a burden, but as the platform that removes complexity and opens the door to everything the cloud native ecosystem has built.
For Akamai in 2026, the mission is concrete: make AI inference workloads great. Through its managed Kubernetes engine (LKE), GPU-backed infrastructure, and distributed edge, the company is building toward a world where developers can run any AI model or workload wherever their users are — without the friction that typically comes with that scale.
