How to Prepare Dataset

Estimated reading: 2 minutes

Select the right conversational topic

Before you start preparing the dataset, you need to choose the right and appropriate topic that you want Kaila to learn and be able to respond on users’ queries.

See the typical use cases of how Kaila can be used to understand better what type of communication or processes can Kaila streamline or fully substitute.

Once you are clear on the task you want Kaila to perform, prepare as much background information as you can for this topic.

Then follow the instructions about how to select training data for AI model training for your use case.

Prepare your training data

After you’ve considered what use case you want to use Kaila for and thought about choosing the right topic for your AI agent, now you need to select and prepare information and materials for training data.

Ideal materials for training data are as follows:

Existing knowledge base content
Product or process manuals
User guides
FAQs
CRM content
Communication channels (Slack, etc.)
Product sheets
Onboarding manuals (user or employees)