Using the Digital Assistant
The Digital Assistant, also known as the chat assistant, serves as the primary interface for interacting with the platform. It's designed to be intuitive and helpful for all your queries.
Getting Started
Navigate to the Digital Assistant module by clicking on the expand/go to digital assistant icon from the AI Innovation Platform Dashboard, from where you can unlock the endless possibilities of our digital assistant.
Starting a New Conversation
To begin a new conversation or a fresh chat session at any time:
-
Click on it to initiate a new conversation and type in your messages or queries in the message box like you would for any other GenAI assistant available.
Providing Context with Documents
You can attach documents to your conversation to provide additional context. This helps the assistant deliver more specific and accurate answers to your queries.
By attaching documents
Look for the attachment icon (paperclip icon) within the chat interface, click the icon to open your file browser and select the desired files from your local device. This would be a session specific context as this attachment would only be added for that particular chat session and would go away in case you start a new chat.
Supported File Formats
The chat assistant supports a wide range of file formats across different categories to accommodate various types of content. Please refer to the table below for a comprehensive list of supported extensions for each category.
Category | Extensions |
---|---|
Document | .pdf , .txt , .csv |
Video | .mp4 , .mpeg , .mov , .avi , .flv , .mpg , .webm , .wmv , .3gpp |
Audio | .mp3 , .mpeg , .wav , .flac , .m4a , .mpga , .mp4 , .opus , .pcm , .aac , .webm |
Image | .png , .jpeg , .jpg , .webp |
Note
When uploading files, you need to ensure they adhere to these supported formats to guarantee successful processing and accessibility.
By grounding the response via existing Libraries
Look for the Library icon within the chat interface, click the icon to open the library and select the desired files from your platform library. The documents in the Library are pre-existing knowledge sources that our digital assistant is integrated with. By selecting any of these libraries, you can ground your queries' response to look for answers in the selected library for better, faster and more relevant responses to your query. More details can be found here.
Navigating Between Assistants
The platform may offer different specialized assistants for various tasks. More details on what these assistants are for and what are they capable of doing can be found here. To switch between assistants, you can follow the below steps:
- Locate the "Assistants" drop-down menu from the chat assistant.
- Select the desired assistant from the list. These are pre-existing assistants available on our platform that can be used to perform various tasks. Refer to this page to understand its capabilities.
Managing Responses
You have several options for managing the assistant's responses:
- Provide Feedback: Help improve the system by providing feedback on the assistant's responses. Look for a feedback icon (e.g., thumbs up/down) next to the response.
- Copy Response: Easily copy the assistant's response to your clipboard.
- Download Response: Download the response for offline use or record-keeping in .txt and .pdf formats.
- Regenerate Response: If the initial response isn't quite what you're looking for, you can ask the assistant to regenerate it.
Advanced Features
Conversation History
You can access your previous conversations through the conversation history panel:
- Navigate to the conversation history section in the left panel.
- Click on any previous conversation to resume or review it.
- Use the search function to find specific conversations by keywords.
Model Selection
Our digital assistant provides you with a wide range of models to select from. You have the flexibility to select the model that best suits your specific needs. Each model is designed to provide answers optimized for different types of requests.
Currently, you can switch between:
- Gemini 1.5 Pro: A powerful and versatile model, ideal for complex tasks, detailed explanations, and creative content generation.
- Gemini 1.5 Flash: An ultra-fast and efficient model, perfect for quick responses, summarizing information, and handling high-volume queries.
We are continuously working to expand our model selection to give you more options. Stay tuned for future releases!
Troubleshooting
Below are some of the common issues faced and their solutions:
File Upload Problems
- Ensure your file format is supported (see table above)
- Check that your file size doesn't exceed platform limits
- Verify your internet connection is stable
Response Quality Issues
- Provide more specific context in your questions
- Attach relevant documents to give the assistant better context
- Try rephrasing your question if the response isn't helpful
Assistant Not Responding
- Check your internet connection
- Refresh the page and try again
- Contact support if the issue persists