Multi-modal Inputs

We're thrilled to announce the launch of Multi-modal Inputs—a powerful new feature that enhances your chatbot's capabilities by enabling file and image uploads during interactions. Now, your chatbot can analyze documents and images to provide insightful, contextually accurate responses based on your training data and base prompts.

What's New

• File & Image Uploads:

  • Activate Multi-modal Inputs in Settings > Chat Interface (available in Standard plan and above).
  • Upload files and images directly during chat interactions.

• Enhanced AI Chat:

  • The chatbot now demonstrates its analytical prowess by examining uploaded documents and images, delivering precise responses to your queries.

• Files RAG (Retrieval Augmented Generation):

  • Uploaded files are processed by extracting and analyzing content to find relevant matches in your knowledge base, ensuring accurate and contextual chatbot responses.

• Images RAG:

  • Two-step Image Processing:
    1. Text Extraction: Extracts text content from uploaded images.
    2. Context Matching: Matches the extracted text against your knowledge base.
  • Customizable Prompts:
    Tailor the text extraction prompt to capture only the relevant details. For example, if your store specializes in specific products, you can use a custom prompt:

    Create concise descriptions of products in the image, including product type, text, etc. Be very concise and straight to the point. Possible product types are:
    - magnet
    - ayna
    - baskili-kese
    - mert-sunnet-davetiye
    - greeting card

  • Test your custom prompt by clicking the Sparkles icon and uploading an image to view the extracted text.

• Live Chat Enhancements:

  • Seamlessly exchange files and images between customers and live agents during chat sessions.
  • If either party is offline, the system automatically sends an email notification with the shared files and images to ensure uninterrupted communication.

Get Started

1. Upgrade: Ensure you're on the Standard plan or above.
2. Activate: Navigate to Settings > Chat Interface to enable Multi-modal Inputs.
3. Customize & Test: Adjust your base prompts as needed, and test image uploads using the Sparkles icon.

Check the Multi-modal Inputs page for more details.