Training
Last updated
Last updated
Trainings allows you to teach and customize your AI agent's responses, ensuring it understands and engages with users like never before. Through the training process, your AI agent learns from example conversations and data, becoming smarter and more accurate in interpreting user intents. Craft personalized interactions, equip your agent with expert product knowledge, and infuse human-like empathy into every conversation
Navigate to the dashboard or workspace.
Select the project in which you want to train the AI agent.
Locate the desired agent within the project and access its settings or configuration.
Once you are in the agent settings, find the Training tab.
Click on the Training tab to access the training feature.
In the Advance Agent Training tab, you will see three training types: "By Example" and "By Rules".
If you need to remove a training example or rule, locate the specific training item.
Select the option to delete or discard it.
You can add multiple training examples to cover various conversation patterns or user intents.
Ensure a diverse range of inputs and desired outputs to train the AI model effectively.
After saving the training examples or rules, it's essential to test their effectiveness.
Initiate conversations or inputs that match the training examples or rule conditions.
Evaluate the AI agent's responses to verify if they align with the desired outputs.
Repeat the testing process with different inputs to validate the training.
Click on the "Inspect" option within the chat bubble.
Once you click on "Inspect," The AI Agent will generate a display showing the relevant training used to generate its response
