Creating Assistant
Last updated
Last updated
Creating an Assistant involves setting up its identity, connecting it to the right data sources, and providing initial guidance so it can deliver meaningful insights.
What to Do: Choose a name that clearly reflects the role or focus of the Assistant, such as “Sales Analyst Bot” or “Customer Insights Assistant.”
Why It Matters: A meaningful name helps users quickly understand the Assistant’s purpose and sets expectations for the type of insights it will provide.
What to Do: Select and connect one or more datasources that the Assistant will access to answer your business questions. This could include sales databases, marketing data feeds, CRM systems, etc.
Why It Matters: By connecting the right datasources, you ensure that the Assistant has the necessary information to generate accurate and relevant insights.
What to Do: Enter guidelines or instructions for how the Assistant should approach answering queries. These instructions might include focus areas, preferred terminology, or specific business rules.
Why It Matters: Clear instructions guide the underlying AI (Language Learning Model) in understanding your business context, resulting in more precise and useful responses.