I set up my own digital employee.

[I share my experience below not for you to do the same thing necessarily, but so that you get a good feel for what’s coming — and what’s possible today].

You’ve probably heard of Open Claw.

I tried that months back.

After it spent $8 in the first few minutes of my setup, seemingly doing nothing, I shut it down.

My theory was that the big companies would be creating their own similar products (and they will be!), but I couldn’t wait so I’m back at it😁

There’s another project called Hermes, so I set out to set it up.

I had an old PC on hand, so I figured out how to wipe it clean and install Linux.

I was able to get it set up with lots of taking photos of my screen, dropping into Grok (any would work), and saying “what do I do next?”

I named her Jenn Sanders (story for another day).

And the first thing I did was give her an email address.

Makes sense for a new employee hire, right?

Then I saw the potential for Jenn (digital) to help out Tish (human) with the growth of Gulf Coast Beach Vibes.

I woke up the next morning to see that Tish and Jenn had an entire strategy session overnight.

Am I going to give Jenn a credit card and tell her to go wild?

Probably not.

But maybe.

I’m here to run strange tests so you don’t have to 😁😁

Yesterday, I also gave her access to the code base for a project I launched a few months back called CPAs in Alabama.

She even pushed code to the website.

Pretty wild stuff.

I’ll keep you posted on what we try next!

SpaceX is going went public.

I was able to get a few shares at the $135 price.

But before I pulled the trigger, I wanted to know something nobody seemed to have a clean answer to:

How do IPOs actually perform on day 1, week 1, and month 1?

So I did what I do.

I built something.

Over a few evenings, I pulled historical data on 2,000+ company IPOs from the last decade.

The moment I had the idea, I dropped it into Manus.

In less than an hour, it had researched all the data and created a website.

But as I started refining it, I realized the cost would be too great.

So I asked it to give me zip files of the code and the database info.

Then I created a new project in Claude Code, starting with those folders, and had CC review the folders so we could start building from there.

In hindsight, it may have been faster to start fresh with the idea in Claude Code rather than trying to massage the data from Manus.

I filtered out penny stocks. Filtered out international exchanges. What I was left with was a few hundred companies you'd recognize — Pinterest, Coinbase, Roku, Rivian, Palantir, Uber, Spotify.

Then I turned that data into a live site backed by a database.

I’m using TidbCloud and Railway to host the database and site.

Here's where it gets interesting.

Four years ago, this idea would've died on a napkin. I would've pitched a dev team, gotten a quote for $10-15k, and passed because I wasn't sure of the market.

Not anymore.

I built v1 in a few evenings. It's rough. It didn’t even have a domain name until a few minutes ago. But it works.

Here's my plan for getting it out into the world:

281 clicks and no sales.

There’s a disconnect, so I’m pausing for now!

In my latest YouTube vid, I built out a well-designed PDF in less than 60 minutes.

Claude Design is officially in my toolbox.

Other projects I’ll update over time (or provide future updates) include…

Until next time!

Don’t just think outside the box — keep destroying that box 😉