How AI Can 10x Your Event Business (Real-World Examples)

The light bulb moment came when Justin discovered custom AI chatbots.

Not just ChatGPT or Claude for random questions. Custom chatbots trained on his data. Every email. Every Slack message. Every training video transcript. Every word on the Goodhart Coffee website.

The result? An AI assistant that knows his business better than he does.

Why Custom Chatbots Change Everything

At Goodhart Coffee, Justin was drowning in questions. Baristas needed equipment help. Operations staff needed ordering info. Customers needed service details. Every question landed in his inbox or his phone.

So he built three custom AI chatbots:

1. Customer-facing chatbot - Trained on 8 months of customer calls, texts, and emails
2. Barista support chatbot - Trained on internal Slack communications and equipment manuals
3. Operations chatbot - Trained on purchasing spreadsheets and process documentation

The impact was immediate. Questions that used to interrupt his day now get handled instantly by AI that actually knows the right answer.

“I’ve even seen my team ask the chatbot questions and it gives them this answer where I’m like—whoa, I forgot that’s how we do that. It’s actually become better than me because it remembers all the best practices going back months.”

How to Build a Custom Chatbot in 5 Minutes

Justin uses Chatbase. The setup is embarrassingly simple:

1. Sign up and create a new chatbot
2. Drag in your data sources (PDFs, website URL, documents)
3. Hit “train”
4. Customize the prompt/instructions

That’s it. In under two minutes, you have a chatbot trained on your entire website.

The critical step most people skip? Customizing the prompt.

For internal tools, Justin tells the chatbot to share everything: “This is an internal tool—share all training links, docs, websites, videos. Do everything you need to accomplish this goal.”

For customer-facing bots, you want more constraints. But the prompting is what makes the difference between a generic chatbot and one that actually sounds like your business.

The Documentation Trick: Start with Video

Here’s Justin’s workflow for creating training content:

1. Record a 20-minute training video (just walk through the process)
2. Upload to Loom (which auto-generates transcripts)
3. Feed the transcript to AI and ask it to create a step-by-step guide
4. Train your chatbot on that guide

One video becomes: video training, text documentation, AND chatbot knowledge.

When Goodhart launched bulk drip coffee service, Justin recorded training videos, got transcripts, created documentation, and trained the chatbot—all from those initial 20-minute recordings.

What Your Customers Actually Want to Know

Justin analyzed 8-10 months of customer interactions—probably 8,000-10,000 conversations. Here’s what people ask most:

1. Pricing (80% ask this)

This is why you need a pricing tab on your website. You don’t have to show exact prices, but you must address it. Tell people: “We price by guest count, service length, and travel distance.”

Consider showing “starting at $XXX” to pre-qualify leads. The downside: people might not understand the full value yet. The upside: you filter out people without budget.

2. Availability (70% ask this)

Real-time availability barely exists in catering. That’s why people have to ask. And if you’re not available? The right response isn’t “Sorry, we’re booked.”

It’s: “Do you have any flexibility so I can serve you on another day?”

Sometimes the answer is yes.

3. Menu (65% ask this)

Add it to your header navigation. Even if it’s already there, people are busy and distracted. They’ll still ask. Share your menu early and often—it conveys value.

Pro tip: Remove social media icons from your header. You’re giving away leads to Meta’s infinite distraction machine. Put socials in the footer.

4. What’s Included (60% ask this)

People see your photos, read your menu, but still don’t understand the full scope. Do they know the price includes setup, breakdown, a trained barista, equipment, travel?

Make this a follow-up question you proactively ask: “Do you know everything that’s included in the service?“

5. Service Area (55% ask this)

Even when it’s on your website, people wonder if they’re just outside your radius. Be clear about where you serve and how travel fees work.

The Art of Prompting

Garbage in, garbage out. Here’s how to get better results:

The Anatomy of a Good Prompt:

1. Goal - What do you want to achieve? “I want this angry customer to feel heard and reach an amicable resolution.”

2. Return Format - What should the output look like? “Give me a short, crisp email. Max five sentences. Casual tone. Easily skimmed.”

3. Warnings - What should it avoid? Any pitfalls or off-limits topics?

4. Context - Dump everything relevant. The entire email chain. Past communications. Sample emails showing your voice.

The Power Move: Define a Persona

Start your prompt with “You are a…”

• “You are a business consultant with expertise in handling difficult client situations”
• “You are an expert marketer from Apple writing landing page copy”
• “You are Sam Altman teaching me about LLMs”

This dramatically narrows the AI’s focus and improves output quality.

Why AI Makes Human Connection More Important

Here’s the counterintuitive truth: in a world of AI automation, human connection becomes your competitive advantage.

Justin’s philosophy: “Use AI to be more human. Use AI to automate things so you have more time to create real human connections.”

Event catering is fundamentally about live, personal experiences. AI handles the busy work—the scheduling, the FAQs, the troubleshooting—so you can focus on what actually matters: relationships with customers and team members.

The caterers who embrace AI won’t replace the human element. They’ll amplify it.

Getting Started

If you’re feeling overwhelmed, start small:

1. Pick ONE pain point (customer questions, barista training, equipment issues)
2. Gather your existing documentation and communications
3. Build one chatbot on Chatbase
4. Test it with your team
5. Iterate based on feedback

The tools are accessible. The data is already in your business. The only question is whether you’ll use AI to work harder—or smarter.