Swix AI: Built So Our Mom Could Use It (And Your Whole Team, Too)
- Jason Swick
- Mar 30
- 6 min read
I've had hundreds of conversations with different travel businesses over the years.
And in almost every single one, the same three problems come up. They're not new problems either, which is the frustrating part. But I think AI is finally in a place where we can actually do something about them.
So let me break down what those problems are, and then I'll walk you through how we built Swix AI to help solve them.
The Three Problems That Keep Coming Up
The capacity problem. This one's probably the most obvious. Expectations keep growing, but the resources don't keep up. There are more channels to manage, more content to create, more personalization that's expected. Same team, same budget. And for a lot of DMOs, hiring isn't always an option. AI is probably the best way to close that gap without burning people out.
The speed problem. Travel moves pretty fast. AI moves even faster. Whether it's a viral moment, a weather event, a new direct flight to your destination, whatever it is, by the time most teams can get around to responding, the moment's already passed. AI doesn't replace good judgment, but it can dramatically compress the time from idea to execution. Instead of a week, maybe now it's hours.
The knowledge problem. I've seen teams where institutional knowledge lives in a dozen different places, and a lot of it is in someone's head. What happens if that person goes on vacation? Or leaves? It feels like cutting off an arm. AI can be that connective tissue. It can hold the context, the history, the brand guidelines, and make that knowledge accessible to everyone who needs it.
Built So Our Mom Could Use It
I joke with my brother Josh about this all the time. Everything we build on this platform should be so easy that our mom could use it.
And if you've spent any time in agentic tools like n8n or Make or some of the other automation platforms out there, you know what I mean. They're fantastic tools. I'm not knocking them. But there's a deep learning curve. And for a DMO team that's already stretched thin, learning one more technical tool is just one more thing on the pile.
We wanted to build something that works for every skill level in an organization.
You're going to have someone at the front desk who just needs to dabble.
You're going to have power users who want to go deep.
And most people are going to fall somewhere in the middle.
Our goal was to make a platform that felt intuitive from day one, but one that people could grow into as their experience evolved.
What's Actually Inside the Platform
Here's how it's set up. Think of it as layers that match where you are in your AI journey.
Conversational AI
Chat. This is the starting point. It'll feel just like what you're used to with Claude or ChatGPT or Gemini. The difference is we've built in the ability to connect to dozens of models and we have guides to help you choose. So if your team likes Claude, great. If someone prefers ChatGPT, that works too. Gemini, Grok, whatever. You can pick the model that fits the task, because certain models are still better at some things than others. You can also connect chat to a local knowledge base and load your own information and context into it. So from the moment you get in, it should feel familiar and useful right away.
AI Teammates. If you've built custom GPTs before, this is similar, but with more muscle. You can give these teammates a persona, connect them to the 3rd party tools your team already uses. We can connect to more than 250 apps. Monday, Google Workspace, Microsoft 365, you name it we probably have it. So your AI teammates aren't just chatting, they're actually working with the systems you work with every day. In my experience, most DMOs haven't really tapped into that kind of capability yet, and that's where a lot of the real power is.
Agents and AI Automations
Level 1- Quick Flows. I think of these as the best place for that two-minute fire drill. You know the one. Your boss needs something at 2 PM on a Friday. Quick Flows let you save your best prompts with user inputs built in. So instead of going back and forth 10 times to get a prompt dialed in every time you need it, you save it once, fill in a couple fields, and you're done. One click. It's the easiest entry point into automation.
Level 2- Action Flows. This is where things start to feel like real AI automation. If you have a process at your organization that goes step one, step two, step three, then output, Action Flows handles that. Maybe the first step collects some information. Then it talks to the LLM. Then it sends something to Gmail. Then it books something on your calendar and sends a confirmation email. It's sequential, it connects to real 3rd party apps, and it can run with or without human review at each step depending on what you need.
Level 3- Smart Flows. This is probably the closest thing to what people mean when they say "AI agents." It's not sequential like Action Flows. Smart Flows can reason, make decisions, and figure out which path to take based on what's happening. Should it go down path A, B, or C? It determines that on its own, maybe loops back, and then checks in with you before moving forward.
What's coming: Agents moving from AI workflows to doing things, monitoring and making choices. This is what's next for us, and honestly, it's the part I'm most excited about. AI that doesn't just follow steps but actually thinks through the problem with you, checks in when it needs to, and gets smarter the more your team uses it. We're building toward that right now, and I think it's going to change how teams work in a pretty big way.
You Don't Have to Build Any of This From Scratch
This is the part that I think really matters for teams that are stretched thin.
Every one of those flow types I just mentioned, Quick Flows, Action Flows, Smart Flows, you can build them manually if you want to. But you can also just tell the platform what you're trying to do in plain English. We call it Generate with AI.
Describe your goal, pick your model, and it builds the whole thing for you. The prompts, the connections, everything. You go in, review it, refine, make sure it's pointed at the right tools, and you're good.
The other thing we've also built is a community marketplace.
DMOs are all trying to solve the same problems. And one of the things I love about this industry is that people are so willing to share what's working. So if someone builds a workflow that solves a problem, they can share it in the marketplace. Another team can download that template, connect their own tools to it, refine as needed and the problem's solved.
The really good automations aren't going to come from us. They're going to come from the DMOs themselves. There are so many smart people in this industry. DMOs all share many of the same challenges and we just want to give them the platform to share what they've figured out so others can benefit as well.
Why It's Built for the Organization, Not Just One Person
Here's something I've noticed in a lot of the businesses I've consulted with. There's always a spectrum.
You've got people like you and me who are trying everything, deep into all of it.
And then you've got folks who typed one prompt into ChatGPT once, didn't like the answer, and haven't been back since. Or they use it only to help them refine their emails or content. That spectrum exists in every workplace whether you like it or not.
The problem I keep seeing is that organizations don't know how to work with all those different groups. Do they just pick a tool for the advanced users and forget about everyone else? That doesn't work.
The way work gets done is going to look very different in one, two, five years. You need a system that meets people where they are and gives them the confidence to grow from level one, to level two, to level 10 and beyond.
That's why Swix is priced for the organization. Everyone's on it.
The people further along can help nurture the ones who aren't. And if someone gets promoted or leaves, all that knowledge stays in the platform. It doesn't walk out the door with them.
A Thought on Getting Started
I think a lot of organizations overcomplicate the AI conversation.
Consultants are great for building a roadmap. But it's kind of like buying a treadmill and never taking it out of the box. The plan only works if you actually use it. To get fit you have to do the work.
If you want a really simple starting point, here it is.
Sit down with your team for an hour. Ask them what processes are just sucking time away from their lives right now. Write those things down.
That's your AI roadmap.
I recently sat down with Tony Carne on the Everything AI in Travel podcast to talk through a lot of this in more detail. If you want to hear the full conversation, you can check it out.
