Dialogflow and Vertex API together is the foundation for building games of the Future.
Working Together:
Dialogflow can leverage Vertex’s capabilities to create more natural and engaging conversations. Vertex can be used within Dialogflow to provide informative and comprehensive responses to user queries.
Bottom Line:
Dialogflow provides the structure for conversations, while Vertex injects intelligence and understanding into those conversations.
TriPilot is a 3-in-1 CoPilot
Specific Context
Model: chat-bison@002
You can send a prompt and the Context that the prompt is to be applied to with Vertex AI Chat-bison@002.
In our Case, we are sending a complete description of game, strategies, point system overview, getting started instruction,…. for the Game to use to answer his or her question about the game.
Example, “Describe the Point System”, “What are my alternatives now”, “How do I see the resources I collected”,….
General Context
Model: text-bison@002
You can send a prompt and request General Context that the prompt is to be applied to with Vertex AI Text-bison@002.
In our Case, we are using Speech to Text from GCP to provide a prompt to the Vertex AI Text-bison@002.
Example, “How did the Boston Red Sox do yesterday?”, “What is the weather in Chicago?”, “Tell me how the stock market is doing today”…..
Database Context
Model: Dialogflow: DialogDeckV3 Agent
You can send a prompt to a PreTrained Dialogflow Agent that can satisfy the request using a Database or other sources of date.
Dialogflow understands conversational sentence structure so you can specify variables to be used in a query.
Example, “is Luke Harrington currently playing on the system?”, “What is the high score for July 2023?”, “What’s my average score over the last month?”,….
Demonstration Video of the new Tri-Pilot