Inspiration
PromptForge was inspired by the need for AI that remembers how users want answers formatted instead of forgetting preferences every chat. It’s built to give users control over the AI’s reasoning and tone.
What it does
It lets users create and apply custom prompt templates that define how AI replies—whether short, detailed, or creative—across all chats and models.
How we built it
We used HTML, CSS, and JS for the frontend, Flask for backend logic, OpenRouter for the AI model, Snowflake API for storing chat history, and ElevenLabs for voice integration.
Challenges we ran into
The biggest hurdles were handling dynamic templates, managing consistent response formatting, and integrating ElevenLabs’ voice feature due to API timing issues.
Accomplishments that we're proud of
We built a working prototype that merges AI customization, data storage, and a clean UI—something that feels like a new layer for all chatbots.
What we learned
We learned how to integrate multiple APIs into one coherent system, store AI data in Snowflake, and balance creativity with technical complexity under time pressure.
What's next for PromptForge
Next, we’ll add PDF and dataset handling, real-time voice input, user accounts, and expand PromptForge into a universal plug-in for any chatbot platform.
Built With
- css
- elevenlabs
- html
- javascript
- openrouter
- python
- snowflake

Log in or sign up for Devpost to join the conversation.