Beyond The Hype: Building Useful AI in Self Storage

Self Storage Lab

Welcome back to Self Storage Lab, the podcast where we deconstruct how operators are adapting to today’s technology-driven landscape—and what it means for the future of the industry.

This episode is a throwback to the early days—with just Mason, Rodolfo, and Kyle at the mics. Together, they unpack where we really are in the AI timeline, why some businesses are pulling back from their GenAI experiments, and how to distinguish between what’s hype and what’s actually working in the real world.

Whether you’re an operator trying to make sense of “AI-powered” products or a team leader wondering how to onboard your staff into an AI-native future, this episode offers both clarity and a reality check.

In this episode, we explore:

• Why 42% of firms have abandoned most of their GenAI pilots—and what’s still worth chasing.
• The difference between AI-powered, AI-native, and agentic platforms (and why it matters).
• Where AI workflows are actually helping self-storage operators today—from collections to customer onboarding.
• How Swivl’s AI architecture blends deterministic and non-deterministic models to do real work—at scale.

If you’re tired of the hype and want the real roadmap for implementing AI in your business, this episode is for you.

In this episode we discussed:

(00:00:20) Blink-182 is old now? Let’s feel that existential crisis together.

(00:01:30) The “trough of disillusionment”—why GenAI pilots are stalling in enterprise.

(00:04:00) Transformers are 8 years old—so what’s really new about today’s AI?

(00:07:30) Why OpenAI, Microsoft, and Google are spending billions on basic research.

(00:09:30) AI isn’t magic—it’s a systems engineering problem.

(00:10:45) Most “AI” is machine learning—and that’s a good thing.

(00:12:30) AI-powered vs. AI-native vs. agentic: breaking down the marketing speak.

(00:16:30) Why consumer expectations are driving product branding more than features.

(00:18:00) Swivl’s 32-model system and what it means to be AI-native.

(00:20:00) Multiverse of agents? What’s real, what’s branding.

(00:23:00) Operator and Rabbit: why AI agents still fall short for real tasks.

(00:25:00) Why Swivl’s AI helps operators—not just consumers.

(00:28:00) OpenAI Connectors—live HubSpot integration and use cases.

(00:30:30) Apple Notes x GPT: the small UX wins that make a big difference.

(00:32:00) What is MCP (Model Context Protocol) and why does it matter?

(00:34:00) Human team analogies for AI agents—building your own AI company org chart.

(00:36:30) Your iPhone camera is the best real-world example of agentic AI.

(00:39:00) Will users ever care how AI works? Nope. They’ll care if it folds laundry.

(00:41:00) Deterministic vs. non-deterministic models: when to use which.

(00:43:00) How Swivl uses both: empathy-driven recommendations + reliable pricing logic.

(00:46:00) Why Apple and Amazon haven’t cracked full AI assistants yet.

(00:49:00) Real talk: banks and airlines would have done this already if it worked.

(00:52:00) What’s next from Swivl—collections agents, onboarding tools, and more.

Mentioned in This Episode:

Internal Tools & Innovations

• Swivl’s stacked lobe system: 32 integrated AI models
• AI-assisted voice agents for collections
• Onboarding agents for new customers
• HubSpot and data platform connectors via ChatGPT

Industry Inspiration

• OpenAI connector architecture
• Apple Notes x GPT integration
• MCP: Model Context Protocol for agent-to-agent comms
• The Economist & IBM reports on GenAI performance

Favorite Nuggets

• AI native = born into a future where GPT is your first study partner
• Deterministic models for billing, availability, and compliance
• Non-deterministic for empathy, tone, and creativity
• Agents aren’t replacing humans—they’re augmenting them with precision

Miscellaneous Mentions

• Sally Sue (our favorite customer persona)
• Baseball analogies and Santa mode GPT
• The joy of talking about octopuses with an LLM

Want to learn more about what AI can do for your facility? Check us out at tryswivl.com or Schedule a demo!