A few of the things we've built, and the ones we talked people out of.
I owned Tour.com and built travel search on it, in an era when that meant writing the index, the crawler, and the front end yourself. There was no stack to pull off a shelf.
The hard part was video. Before broadband it barely existed on the internet, and tourism sells on video. Meanwhile a printed price is wrong a week after you print it. So I designed a system where the heavy lifting sat on a CD-ROM, one per city, video and all, and only the volatile data came over the wire. Pricing, availability, whatever a business changed that morning, synced down over a slow connection in real time. Tour.com went to a bidding war and sold.
The video skills from Tour.com turned into a second business. I produced training and procedure videos for the Louisiana Department of Health, and then for several pharmaceutical companies: onboarding, compliance, and product training for field reps and clinical staff.
It's a discipline that looks nothing like software and turns out to be the same job. Understand what the person on the other end actually needs to do, strip everything else out, and check that they can do it afterward.
A client was running their operation on cloud services with internet that dropped constantly. Warehouses in places where the connection was never going to be reliable, and a business that couldn't stop when it went down. So we built synchronization between their local servers and the cloud. When the line failed, everything piggybacked onto local and the floor never noticed. When it came back, it re-synced. Years later the industry gave that idea a name.
I designed a route management system for tracking students on and off school buses by RFID. Patent pending.
A growing share of our work is taking something a founder built with an AI tool over a weekend and getting it through its first real month. The prototype is usually further along than they think. The database, the permissions, and the payment flow are usually further behind than they think. We fix the second list without throwing away the first.
I've told a number of companies that the AI tools they were looking at would do very little for them at the expense being quoted. That advice is worth more than most of what gets sold, and I'd rather give it and lose the sale.
We'll look at your business first and tell you honestly whether we're the right fit for it. That conversation doesn't cost anything.