By MCP servers integrated into enterprise applications are introducing latency that stalls AI agent workflows and forces every downstream operation to wait. Developers building long-running tasks, in-agent forms, or multi-step automation hit this wall as a hard blocker. The result is degraded user experience and agents that cannot run at production scale.
In this interview on TFiR, Du’An Lightfoot, Sr. Developer Advocate for AI Engineering at Akamai, breaks down how the latest MCP specification enables async execution, what that unlocks for enterprise agent workflows, and what is still missing from the spec.
Guest: Du’An Lightfoot, Senior Developer Advocate for AI Engineering at Akamai
Show: The Agentic Enterprise
LinkedIn: Du’An Lightfoot
Watch the discussion on YouTube
Here is what every AI platform engineer and developer building agent workflows needs to know.
Technical Deep Dive
Q: Why does MCP latency block AI agent workflows in enterprise applications?
Du’An Lightfoot, Sr. Developer Advocate for AI Engineering at Akamai, explains that MCP servers connected to enterprise applications create a wait-state problem: when an AI agent requests information or triggers a task, the application has to hold until the MCP job finishes before it can proceed. For developers building agents that handle long-running tasks, forms, or multi-step processes, this blocking behavior is a direct performance bottleneck. The impact shows up as increased response times and a worse end-user experience.
“The problem that I was having was that I had this hangup with MCP. It was adding latency to my application.”
Du’An Lightfoot, Sr. Developer Advocate – AI Engineering, Akamai
Q: How does async MCP execution reduce latency for AI agent applications?
With async MCP support, an application no longer has to wait for a job to complete before moving forward. Lightfoot describes the shift as a direct lever for reducing perceived latency and improving user experience: the agent kicks off the MCP task, then continues processing other parts of the workflow while the result resolves in the background. This is especially significant for long-running tasks and workflows that previously had no way to decouple request from response.
“Rather than waiting for this job to finish, you could move on to other tasks within your application, potentially reducing latency and improving the user experience.”
Du’An Lightfoot, Sr. Developer Advocate – AI Engineering, Akamai
Q: What new use cases does the updated MCP specification unlock for developers?
Lightfoot points to long-running tasks, in-agent forms, and full applications running inside MCP as categories that become practical with the new specification. Developers who previously could not build these workflows because blocking MCP calls made them unusable now have a path forward. He frames this as an opportunity to revisit workflows that were shelved because the spec was not ready, not because the use case was wrong.
“This is an opportunity to figure out: I’ve been wanting to build these specific workflows, but the problem I was facing was that I had this hang-up with MCP.”
Du’An Lightfoot, Sr. Developer Advocate – AI Engineering, Akamai
Q: How should teams approach MCP adoption at scale given how new the specification is?
Lightfoot is direct that because async MCP is new, there is no established playbook yet. Teams need to experiment: run the workflows, find what works, find what breaks, and learn how async behavior interacts with their specific application architecture before rolling it out broadly. He does not present a prescriptive framework, but emphasizes that experimentation at small scale before enterprise-wide rollout is the practical path.
“Because it’s so new, it’s going to take some experimentation to figure out what works, what doesn’t, and how to actually apply it at scale within your organization.”
Du’An Lightfoot, Sr. Developer Advocate – AI Engineering, Akamai
Q: What is missing from the current MCP specification and what should the next iteration include?
Lightfoot’s top wish-list item is a centralized MCP server repository, modeled on an app store. His reasoning: as agents become common on phones and in enterprise environments, developers and users need a single place to discover, download, and connect to MCP servers rather than managing integrations file by file. He sees MCP as a future standard for app integration more broadly, and a discoverable repository as the missing infrastructure layer that would make that vision real.
“I would like to see a central repository of MCP servers in the future, much like an app store, where everyone can go, download an MCP server, connect to it, and access the data and tools they need.”
Du’An Lightfoot, Sr. Developer Advocate – AI Engineering, Akamai
Q: How does MCP compare to skills-based agent integration for app developers?
Lightfoot compares MCP to skills-based integration models used by agent frameworks such as Hermes and OpenClaw. Skills are one path for app developers to connect their applications to agents, typically by downloading and bundling a file. Lightfoot’s position is that MCP offers a cleaner alternative: connect directly to the MCP server instead of distributing a file, reducing friction for both the app developer and the agent user.
“Why download a file when you can simply connect to this MCP server?”
Du’An Lightfoot, Sr. Developer Advocate – AI Engineering, Akamai
Resources & Documentation
- Akamai, cloud and edge platform with AI engineering and developer advocacy resources for agent-based architectures
- Model Context Protocol (MCP) specification, the open standard for connecting AI agents to tools and data sources
***
👇 Click to Read Full Raw Transcript
Swapnil Bhartiya: If you just look at from purely developer’s point of view, how he’s going to solve some of their problems, some of their struggles, some of their frustration, when their day to day life has changed. They got access to the things they want, they got things the way they want and also security wise because it is actually every time you plug into, you have to think about so many things and sometimes you cannot think about that. So talk about the impact on developers day to day life, who is totally hooked into mcp?
Du’An Lightfoot: Yeah, from a developer perspective, a lot of developers are using MCP servers. They’ve built their own MCP servers, depending on how deep they are in AI. When it comes to this migration, it’s a question of, okay, does this MCP server solve my needs? Right. If you have one in your organization and it connects to your application. Well, now if you’re building AI agents that request information, you can do a lot more, right? Long running tasks, forms, apps within mcp. This is an opportunity to figure out, okay, I’ve been wanting to build these specific workflows, but the problem that I was having was that I had this hangup with mcp. It was adding latency to my application. Well, now what you could actually do is rather than waiting on this job to finish, you could actually move on within your application, maybe decrease some latency, increase user experience. There’s a lot of different things that you can do from a developer perspective, but I think because it’s so new, it’s going to take some experimenting and finding out what works, what doesn’t work and how to actually apply it at scale within your organization.
Swapnil Bhartiya: What are other things that you wish should have made into this specification or that is on your own wish list for the next iteration of specifications.
Du’An Lightfoot: When it comes to wish list, I see MCP as being the future of maybe app integration, right. Since we have tools like Hermes, Agent, openclaw and they’re using skills. But if I am an app developer and I want to integrate with these agents, skills are one way to do that. But why are we downloading a file when you can just connect to this MCP server? So I would like to see in the future if it’s possible, if it happens, if there is maybe like a central repository of MCP server, like an app store where everybody that has, which I believe they will, an agent on their phone can go to the app store, download whatever MCP server, connect to that MCP server, get the data, get the tools that they need so we can all build new things and be more efficient, not just in their enterprise, but in life.
