Some notes from Missouri Startup Weekend 2026 (MOSX)

72 pitches, one final winner - mega Shark Tank.

Apr 23, 2026

I came home from Missouri Startup Weekend, oddly energized with a full Apple Notes app.

What is Startup Weekend?

For those who haven’t been to one…

Startup Weekend is a 54-hour event where strangers form teams around ideas pitched on Friday night, build something over the weekend, and present to judges on Sunday.

It’s part hackathon, part business plan competition, part crash course in entrepreneurship.

The format has been running for over a decade, but what people build in those 54 hours keeps evolving with the tools available.

This year’s numbers

Friday night opened with 72 pitches — one minute each to sell an idea to the room. Judges and audience voting narrowed that down to 15 teams that would actually form and build over the weekend.

By Sunday’s presentations, there was almost $140,000 in real investment commitments made to teams that had existed for less than three days.

This year’s Missouri Startup Weekend drew participants from across the state — students, working professionals, serial entrepreneurs, and plenty of people trying their first startup.

Before the energy of all this excitement away, I wanted to jot down a few thoughts from the weekend.

Lots of conversations about AI, as expected

I got to talk to ALOT of people building with AI, or using AI in some shape, way, or form.

That alone feels widely different from even a year ago.

Autonomous bulldozer??

One of the most memorable moments on the hardware side was at EquipmentShare’s robotics lab.

They let us remotely operate a robotic bulldozer from their technology center, with the actual machine sitting in a barn in Fulton.

The same bulldozer also accepts voice commands from the phone that you can probably just barely see above the Xbox controller. It can also move autonomously within the space.

Another interesting feature is that if you look really closely at these images you can see cones in the middle of the screen. If you try to drive the bulldozer into the cones, a ‘shine detection’ sensor will detect the cones and stop the bulldozer.

Let’s say a worker on the job site falls over while wearing a shiny protective vest; this feature helps the robotic bulldozer avoid running over the worker.

As you may expecty, safety is a big concern when rolling something like this out in real life.

Here is my friend Caleb picking up some dirt with his new bulldozer

Adoption is faster than it’s ever been

I stopped by to see my friend and student Andrew, who was already designing a MVP for the AirWaze in Claude Design.

That specific Claude Design feature had launched roughly 48 hours earlier — Anthropic released it on April 17 — and here it was, already in a student’s workflow at a weekend event in Missouri.

That’s a data point I keep coming back to.

The lag between “launched at a frontier Silicon Valley AI lab” and “a college student in the Midwest is shipping with it” used to be measured in months or years.

This weekend, it was measured in days. Whatever you think about the pace of AI itself, the pace of adoption around it is now the more striking number.

That’s overwhelming, but also kind of cool.

Thoughts from the top pitches

As the creator of a card game designed to help you pitch your ideas better, you bet I had thoughts on what teams could have done better, new trends, and what they did well.

Getting ready for the team’s final pitches

A few of top teans — quick impressions

A handful of the teams that placed, with the one-liner I scribbled down for each:

Tonetta — coaches people to speak more clearly on sales calls
Unileaae — student housing search, because finding it currently takes forever
EQUIPR — lab equipment rental; they pointed at a roughly $1B market
CMPLX — Zillow for off-market real estate deals
CRANK — student organization monetization and event discovery; the team itself had more energy than anyone else in the room
SBA Loans App — an agent to help small businesses through SBA loans; ~$33B in SBA loans annually

Some of my favorites

The ones I thought had the sharpest shot:

RIVE — a better way to do health savings accounts - cut out the middle man, and see if the consumer comes out better.
Airwaze — Helping with manual flight loggined
AI Dungeon Masters — AI to help run DnD games aimed at the ~30M D&D players
HerdRoute — Uber for farm animals, connecting auction houses, buyers, and drivers. Genuinely novel, and the kind of unsexy logistics problem that real businesses get built on

I wouldn’t recommend using Claude for presentations just yet.

Since I was just singing praises of the speed of adoption, I do one to give one honest note about Claude for pitches like this.

For actual pitch decks and presentations, I would not use Claude just yet.

Some teams presented fully Claude-made presentations. They were super difficult to read and understand. They were very text-heavy, making them almost impossible to digest.

Instead of being a visual aid, they probably actually hurt their presentations

Demos of functional MVPs

The single biggest change from past startup weekends is that almost every team had at least an MVP.

For non-technical folks, that is a minimum viable product - something visual you could actually interact with or watch run.

So, we are talking not a sketch on a napkin, not a Figma mockup, and definitely not three screenshots taped to a Google Slide.

Pretty well-designed working demos.

Tools like Claude, Cursor, v0, and now Claude Design have collapsed the distance between “idea” and “something I can show you.”

54 hours used to be barely enough to nail down a concept and get a mock-up of what your idea was.

Today, it’s enough to ship a rough but pretty functional, nice-looking version.

I talk more about the speed of software development today in my newsletter: Everyone Gets a Jarvis - the build versus buy math,

What the best pitches had in common

The strongest pitches of the weekend all did one thing really wel,

They incorporated direct customer feedback, and they quoted customers by name.

One of the very best teams went a step further and played short video clips of real customers discussing the validity problem.

You can’t fake that, and you can feel the difference the second it hits the room.

If I had one piece of advice for next year’s teams, it would be to spend less time on your slide design and more time getting on the phone with five potential customers.

Then put them on the screen and talk about how they love what you are doing. It sells really well.

That’s my weekend.

Fulton bulldozer, a student building in a two-day-old AI tool, and a field of teams that mostly showed up with something that actually worked. Good weekend to be in Missouri.

Discussion about this post

Ready for more?