Much appreciated @bubble.trouble!
Forgot the about page and pitch:
About page: StoryPilot - You lead the way! (www.storypilt.io/about)
Old pitch for TechBBQ in CPH 2024: https://app.box.com/s/7jyunftcyfhmelspog7cezfhyze0jd4s
I am aware that POI generation is slow. It fetches everything on the fly. POIs from all wikipedias (various languages) from a pool of 40M + places, then generates the markers with a CDN (content delivery network). Once you press the POI, text is streamed from OpenAI, then narrated using OpenAI as well.
Data pool
Most other apps have a limited library of pre-written and generated POIs. The idea with this app is to infinite data. Even the story text are adjusted to the narrator characteristics on the fly and there are loads of possibilities to improve them. Eg. remember the last 20 stories and make them “flow”… not be like the previous and change small details.
Speed improvements
The speed can be improved, both upgraded to a newer micro GPT model, the CDN can be speeded up or better. Generate the markers with a js library in the APP instead of using an API - This is by far the biggest bottle neck and will improve the speed considerably.
Also, marker resolution can be lowered and POIs pre-generated based on app usage and location. At the moment POIs are stored once generated. So… If you use the app a lot in a local city, you will have most of the POIs generated after 2-5 app opens, depending on the city size. Once in the database, they lost almost instantly.
Also, OpenAIs streaming TTS is slower than eg. Play.ht. Which also has emotional control. Its quite straight forward to make the swap with the plugins.
The concept
- hands free. Not more “finding places”. Hit play and go. You focus on the real world
- inspired by the pandemic isolation, and going off the beaten path, POIs are not “just” the typical places, but all the things you didn’t know existed. That was the idea to begin with.
- Custom narration. No more cached generic stuff. Its made for YOU.
Monetisation
Okay, it will take time. But at least you have a tested and almost finished product. I made the app in close collaboration with the truism industry orgs. in Denmark. Several business devs. was involved. About 400 hours spent. Little for getting this far, but it was fun.
Several firms showed interested:
- Hotels: Offering it for free, collecting user emails for events and conferences. 1-2 USD pr mail
- Scooter rental: Incorporating it for free in their model. Classic licencing model
- Paywall: let the user pay a small subscription price after 5 stories listened. Eg. 9 USD pr month. Go for mass market and be the international-worldwide alternative to the hordes of sub-par guided audio tour apps.
- Sell advertisements: Add POIs for restaurants and other places. Let the pay a small price to customize their POT story / text.
Other info
- Cost about 100 USD pr month to run
- Comes with iOS and Android apps - Try them out. They have “Autoplay” function that reads stories once you pass them. Hands free. Wupwup.
If you have start-up / scaleup experience and 5-10 hours a week to spend. Or a Bubble dev to pay a bit for UX / speed improvements. This will give a decent head start… Or better. Affiliated with a Hotel or hotel chain. Let them offer it for free on their counters with a QR code
Charge them a monthly fee.
Or spend 10-20 hours on turning it into a template on the marketplace and see what happens
I earned 30K USD on similar apps, but much less tightly integrated and less functionality.
@boston85719 Remember to turn on sound. Download the iOS or Android app for the full experience. The web app cant fetch your background location and thus, not auto play stories when you get near them. The native apps better.