By Most distributed AI pilots stall or fail outright when they move toward production. Teams build agents that solve no clear business problem, ship without load validation, and architect without defined SLOs, then discover the gap only when systems fail under real-world conditions.
In this interview on TFiR, Jon Alexander, SVP of Product, Cloud Technology Group at Akamai, walks through the crawl-walk-run framework enterprises need to move distributed AI from experimentation to production-grade deployment, covering guardrails, scalability, security, and SLO-driven architecture.
Guest: Jon Alexander, SVP of Product, Cloud Technology Group at Akamai
Show: The Agentic Enterprise
LinkedIn: Jon Alexander
Watch the discussion on YouTube
Here is what every platform engineer and AI infrastructure team needs to know.
Technical Deep Dive
Q: How should enterprises structure the move from distributed AI experimentation to production deployment?
Jon Alexander, SVP of Product, Cloud Technology Group at Akamai, frames the journey as a deliberate crawl-walk-run progression. The crawl phase focuses on confirming the agent solves a real business problem and that the right guardrails and controls are in place before any further investment. Skipping this step is the primary reason AI agents get built but never deliver value.
“Focus on whether you’re solving a real problem. Many agents fail to deliver value, so make sure you’re addressing a genuine business problem.”
Jon Alexander, SVP of Product, Cloud Technology Group, Akamai
Q: What does the walk phase of distributed AI scaling actually require?
Alexander describes the walk phase as the point where teams must verify the system can handle realistic production load, is ready for real-world deployment, and is secured against the threats that come with exposure at scale. Each of these is a gate, not a checkbox, and failure to clear any one of them will surface in production.
“Make sure it can handle the expected workload, is ready for real-world deployment, and is secure.”
Jon Alexander, SVP of Product, Cloud Technology Group, Akamai
Q: How do SLOs and performance benchmarks determine whether an AI system is production-ready?
The run phase is where Alexander says teams must architect specifically to hit defined SLOs. For physical AI, autonomous vehicles, and commerce use cases, end users and dependent systems carry hard expectations. If the architecture was not designed to meet those targets, the system will fail its business objectives regardless of how well it performed in testing.
“Make sure you have clear performance benchmarks to target and that your architecture is designed to meet them.”
Jon Alexander, SVP of Product, Cloud Technology Group, Akamai
Q: Where does Akamai fit in scaling distributed AI performance for production workloads?
Alexander positions Akamai as the layer that helps enterprises scale out and scale up performance once the architecture and SLO targets are established. The focus is on meeting the performance expectations that real-world distributed AI execution demands, particularly at the edge where latency directly determines whether the system succeeds.
“That’s where Akamai can help you scale out and scale up performance to meet those objectives.”
Jon Alexander, SVP of Product, Cloud Technology Group, Akamai
Resources & Documentation
- Akamai Blog, updates on edge AI infrastructure, performance scaling, and distributed cloud evolution
- Akamai Cloud Technology Group, distributed cloud platform for scaling AI workloads at the edge
***
👇 Click to Read Full Raw Transcript
Swapnil Bhartiya: Infrastructure providers like Akamai continue to evolve. How should enterprises rethink their own architecture to prepare for the shift towards distributed AI execution?
Jon Alexander: I think it’s a cruel walk run. I mean, typically what we’re seeing is people are experimenting, they’re building the application, and I think one of the key things is focus on are you solving a real problem? Like, there’s a lot of agents that don’t end up delivering value and, and therefore confirm that you’re really solving a business problem. Make sure you’ve got the right guardrails, make sure you’ve got the right controls in place. That’s the first thing. So that’s the crawl phase. Then the walk phase is like, hey, how are you going to scale this out? Make sure that this is able to handle the load that’s coming, make sure it’s ready for real world deployment, make sure it’s secure, like that key. So that’s the walk phase. The run phase though, is where you get into the optimization around performance and make sure it’s meeting the expectations. Again, back to some of the examples we talked about. So for physical AI, for an autonomous vehicle, even for commerce, there are expectations that end users have or that systems have, and if you’re not meeting those expectations, the system’s gonna fail, it’s not gonna meet the business objectives. And so that’s the final level of like, hey, make sure that you’re actually achieving the SLOs. Make sure you’ve got really clear performance monitoring benchmarks that you’re trying to target and that you’ve architected to achieve those. And that’s really where Akamai can help you scale out and scale up the performance to meet those objectives.
Swapnil Bhartiya: John, thank you so much for joining us today and sharing these insights on latency and what it really means for the future of AI infrastructure in production. For everyone else watching, please make sure to check out Akamai and their blog to see how this market is evolving and how you should be preparing. John, thank you so much for your time and I look forward to chat with you again.
Jon Alexander: Great. Thank you very much. Cheers.
