/simulai

simulai is a Notion-inspired open-source and free conversational survey builder, powered by AI.

Primary LanguageVueGNU Affero General Public License v3.0AGPL-3.0

🤖 simulai

simulai is a Notion-inspired open-source and free conversational surveys builder, powered by AI.

Effortlessly design your interactive forms that mimic natural conversations, powered by AI. Let AI answer users inquires regarding your questions/survey/form within conversational forms. Ideal for gathering insights and engaging users in interactive surveys, our tool is inspired by the simplicity of Notion and is completely free and open-source.

simulai-demo.mp4

Features

  • Block-based chatbot editor
  • Conversational forms powered by AI (connect your API: OpenAI, Claude, Llama AI, ...)
  • Provide data context for AI to answer users inquires in detail
  • Drag to reorder blocks
  • Basic Markdown-parsing including bold, italic, URLs, headings and divider
  • Type '/' for command menu and shortcuts

Contributing

1. Clone this repository, go to the root directory and install packages

Please use node >= v16.15.1, python >= 3.10.1

git clone https://github.com/simulaiofficial/simulai
cd simulai
npm i
virtualenv venv
pip install -r requirements.txt

2. Run dev

uvicorn api.main:app --reload --port 5678 --host 0.0.0.0
npm run dev

2. Run prod

gunicorn api.main:app --workers 4 --worker-class uvicorn.workers.UvicornWorker --bind 0.0.0.0:5678 --daemon
npm run build
npm run preview

If you head to http://127.0.0.1:5173/?page=http%3A//127.0.0.1%3A5678/page%3Fsrc%3Dhttp%3A%2F%2F127.0.0.1%3A5678%2Fsrc/1234 on your browser, you should see what looks like the screenshot above.

3. Contribute!

Simulai is quite limited for now but we hope it serves as a good starting point for other folks looking to build their own chatbot editors.

We would love to make Simulai more extensible and welcome any suggestions or contributions!

See CONTRIBUTING.md for details.

Acknowledgements

Initial code was based on Lotion repo built by Dashibase, thank you!

This was made much easier with the following libraries and frameworks, thank you!