Thilo Huellmann

Co-founder @
Hey thereΒ  πŸ‘‹

as promised in my last post. Here is the full, step-by-step tutorial πŸ‘‡

T30 Project:

Let me know if you're interested in an invite to colabel's private beta :)

I'll publish the bubble template soon πŸš€

HiΒ  Thilo Huellmann Β 

I enjoyed the walkthrough thank you. I can see the power of colabel and it is impressive!

If I were to train a model with Spanish text I do not see how the model could not work. Would it?
Hi Carlos Marchan Β 

thank you! Yes, Spanish text is also possible. You can contact me via to get access :)
Hey everyone πŸ‘‹

just submitted my project for the T30 challenge:

Problem πŸ€”
I realized that millions of people use StackOverflow but only very few ask questions. Especially beginners are hesitant to ask any new questions.


Because the community on StackOverflow has no mercy with inexperienced people. Many questions are downvoted or closed without comments. Not a great experience! :(

My solution to this problem πŸ’‘

I want to give people a "safe space" to check their questions quality before posting it. I found an amazing dataset on Kaggle that consists of 60k historic StackOverflow questions + their rating. I used the data to train a text AI with colabel that can predict the quality of any new question with 90%+ accuracy.

My no-code stack πŸ’»

I only used bubble (for the web app) and colabel (for the AI) and connected it via a simple plugin. It was the first time for me using bubble after some learning I felt really powerful with it (still have to get used to responsive design though).

I will also publish a video of the full building process later today :)

You can find the app here:

Q: Should I make a bubble template out of this? :)
AWESOME! Def make a Bubble template!