📝Article2Rasa🤖
Article2Rasa: A simple tool that generates RASA format NLU from any article, enabling users to create custom chatbots within minutes effortlessly. No NLP expertise or YAML file writing required.
For developers and maintainers, it offers a trimmed version of RASA 2.8.2, offering ease of scalability, utilization, and maintenance, simplifying the entire process.
📚 Demo Links
📖 Story Block
I am actively involved in the dynamic world of e-commerce, catering to multiple clients who own their thriving malls.
Despite having a wealth of valuable FAQ articles to enhance customer support, these clients currently lack expertise in NLU and RASA technology.
To unlock their full potential and foster healthy competition, we have the perfect solution: introducing the revolutionary Article2Rasa tool.
This user-friendly tool empowers users to build their own chatbots quickly, delivering an unparalleled customer experience that will set them apart from the rest of the market.
🤔 Comparisoon
LLMs may provide inaccurate answers, making them might be unsuitable for the e-commerce industry.
Budget constraints further complicate their use.
ChatGPT / LLMs
| Pros | Cons |
|---|---|
| Automatic generation of rich responses | May have incorrect knowledge |
| Versatility | Reasonableness and richness of content generation |
| Dialogue context management | Slower Response speed |
| Learning capability | Development, training, and deployment costs |
RASA NLU
| Pros | Cons |
|---|---|
| Faster Response speed | Flexibility in content generation |
| Stability | Limitations in versatility |
| Lower development, training, and deployment costs | Difficulties in managing dialogue context |
🔧 Features
- User-friendly interface for effortless modification of FAQ articles
- Creation of RASA NLU from FAQ articles using ChatGPT or LLMs (Language Model)
- Automatic fix YAML and RASA format errors
- Straightforward RASA server for convenient usage
💻 QuickStart
- Checkout quick_s_tart.md
🐛 BUG / TODO List
- Sometimes LLM's output can't be parse properly
- More customization options (e.g. language, input format, etc.)