Dialogflow provides features that can help you build and refine your agents using real world data. As a developer, you can leverage existing sources of conversation data you might have access to, as well as usage and performance data that pertains to your running agents.
There are three main use cases for real world data within Dialogflow:
- Building a new Dialogflow agent using logs of existing customer interactions.
- Use training to add existing data to training phrases.
- Improving the performance of a live Dialogflow agent using its own logs.
- Use training to add additional data to training phrases.
- Understanding the performance of a live Dialogflow agent to inform your design decisions.
- Use analytics to assess performance of agent.
The documents in this section describe Dialogflow's training feature and how you can use training to do the following:
- Build and improve agents using customer interaction logs.
- Upload logs from an external source.
- Incorporate logs from a live agent.
- Get insight into how well your agent is performing.