HN user

hafiz_

7 karma
Posts4
Comments7
View on HN

I built this because I kept getting mediocre AI responses, turns out my prompts were the problem, not the AI.

The two stage thing came from Reddit feedback. People were abandoning the tool when I asked questions upfront, so now it gives you a quick result first, then optionally asks questions for a better version.

Stack: Laravel + Vue.js + OpenAI/Deepseek API

Been running for a week now - 207 users, 1.7K optimizations, which is cool. Mostly trying to figure out how to keep it free without going broke on API costs (currently limited to 25/day per user).

Would love feedback on: - Does the two stage flow make sense? - What features would actually be useful? - Are 25 optimizations/day reasonable, or too stingy?

Happy to answer any technical questions.

Thank you for the valuable feedback. I understand your points about the UI and the speed of interaction. I am considering different ways to enhance the user experience, including possibly a quicker color selection method and a more interactive canvas like feature. Your suggestion about using a predefined palette is a great idea, i will look into implementing that to improve speed and user engagement. Thanks for helping me make this better ...

Thanks for the great point. Currently, i don't have a validation system for the names submitted by users, but I appreciate your concern and see the potential issues it could pose. I'm planning to adjust the submission process. Users will only submit the hex code, which is straightforward to validate, and then i will utilize the 'webcolors' Python library to automatically assign a standardized name based on the hex code. This should help maintain consistency and prevent any inappropriate naming. Thanks again for your valuable feedback