Before your agent can answer questions, you’ll need to ground it with knowledge. This will help it understand your specific information and answer questions more accurately.
Upload relevant documents (PDFs, text files, Google Drive documents, etc.)
Add website URLs for the agent to learn from
Link YouTube videos that contain relevant informations or tutorials
Behavior: Define the agent’s role, personality, and communication style. Be specific about how the agent should interact. Use the built-in prompt generator to create and refine your agent’s instructions.
Model: Select the LLM powering your agent (e.g., GPT-4, GPT-4o mini). Each model consumes different amounts of tokens and has different capabilities.
Knowledge Restrictions: Choose whether to restrict the agent’s responses to only your provided knowledge content. Make sure to specify to the agent what to answer if the knowledge base doesn’t contain an answer.
Emoji Support: Enable or disable emoji in the agent’s responses.