Machine Speed Attacks, Voice Agents, and Why Bad AI Excuses Fail with Keith Trawick - Ep 218

About this episode

Guest Introduction

Keith Trawick is the CIO of Stretch Zone, a practitioner-assisted stretching franchise with more than 420 locations across the country and another 75 to 80 expected to open this year. He joined the organization as employee number one when it made the move from boutique wellness service to scalable franchise brand roughly 12 years ago, helping build the technology infrastructure from the ground up in a category that did not exist before Stretch Zone created it. With a career rooted in subscription-based, member-centric businesses, Keith brings a systems-first perspective to the intersection of AI adoption, franchise operations, and the very human challenge of bringing hundreds of independent business owners along on the journey.

Here's a Glimpse of What You'll Learn

• What it means to build technology for a franchise category that did not exist when you started and how that shapes the systems-first philosophy Keith still operates from today
• Why Keith believes the service Stretch Zone delivers is AI-resilient at the front line and where the real AI opportunity lives on the back end
• How machine learning in security tools is the unsung hero of the current threat environment and why traditional patching alone cannot keep pace with machine-speed attacks
• Why Keith is deploying AI voice agents for inbound and outbound calls across the franchise network and the data foundation problem that has to be solved first
• How he is partnering with SoundHound on voice and Blend on middleware to build an agentic system that respects compliance requirements across hundreds of independently owned locations
• Why outcome-based pricing for AI tools makes more sense than hourly labor for a franchise model and what that calculation looks like in practice
• Why the organizations that wrote off AI after a bad ChatGPT hallucination experience are going to have a very hard time competing from here

In This Episode

Keith opens with an origin story that reframes what technology leadership looks like when you are building the category, not just the company. Stretch Zone did not have a Google Business Profile category to select when it launched because no such category existed. Nobody knew what getting stretched meant. Keith joined as employee one with the franchise growth model and has spent 12 years building the systems infrastructure that allows more than 420 independently owned locations to deliver a consistent, brand-defined member experience without micromanaging the owners running those businesses. That tension, between brand consistency and franchise autonomy, runs through every technology decision he makes, and it is the lens through which he evaluates every AI initiative the organization is now pursuing.

The security section of this episode is where Keith gets most animated, and with good reason. He draws the machine learning versus LLM distinction with a water-in-the-boat analogy that lands harder than most technical explanations do. Traditional patching is reactive by design: the boat manufacturer notifies you of a defect, you patch the hull, done. But zero-day vulnerabilities exploited at machine speed do not wait for the notification cycle. What Keith wants is a system that detects water in the boat as it arrives, identifies where it came from, and addresses it before the hole is officially documented. That is what machine learning tools like Darktrace are doing in practice, and Keith makes a direct case that behavioral AI understanding what is normal for each user, each application, each network pattern, and flagging deviation from that is the only defensive posture that makes sense when the attack pace has crossed from human speed to machine speed.

The voice agent initiative is the most concrete and forward-looking section of the episode. Keith is mid-implementation, weeks away from beta testing at targeted locations, and he is candid about exactly how complicated it is to deploy agentic AI responsibly across a franchise network. The technology problem, building an AI that can handle inbound member calls and make outbound follow-ups within the right guardrails, required choosing partners with deep expertise rather than assembling something from YouTube tutorials and automation harnesses. SoundHound handles the voice side. Blend handles the middleware and data layer. But what took the most work was building the data foundation underneath it: a consistent definition of what a member actually is across 420 locations where 100 different owners might give 250 different answers to that question. Keith is clear that the agentic capability is ready. The last mile is compliance, making sure outbound call campaigns are registered, approved, and respectful of each state's quiet period rules across hundreds of independently owned businesses. That is the problem he is solving in real time, and the fact that he is talking about it before the rollout rather than after makes this episode particularly valuable for anyone who is contemplating the same move.

This episode is brought to you by Cyberlynx