At the heart of Preemptive AI for Domino is the AI-Proxy database.
This database stores configuration information for Instructions (such as "Ask") and specifies the Service that should process each instruction (for example, LM Studio or Domino IQ).
This document explains how to configure Preemptive AI to use Domino IQ as its LLM server.
Prerequisites:
You should already know how to configure Domino IQ. If you need help, please refer to HCL’s documentation or contact us by creating a support ticket.
Configuring Preemptive AI to Use Domino IQ
Preemptive AI communicates with Domino IQ using a single System Prompt and a single Command. In this example, we’ll use the display name "Preemptive-AI" for both.
1. Set Up the System Prompt in Domino IQ
- Open the Domino IQ configuration database.
- Select System Prompts and add a new System Prompt. Make sure you enter the text: Process the user request
- Fill in the form as shown below and save it.
2. Set Up the Command in Domino IQ
- Select Command and add a new command. Make sure the Command and System Prompt name all match up.
- Fill in the form as shown below and save it.
Make sure to select a Configuration and Model that you have already set up and verified as working.
You have now defined the "Preemptive-AI" System Prompt and Command.
3. Configure the Domino IQ Service in the AI-Proxy Database
- Open the AI-Proxy database.
- Select Services and edit the Domino IQ service document. Alternatively, you can create a new service document with a different name.
- In this document, ensure the following:
- Processor is set to Domino IQ.
- Domino IQ Server is set correctly.
- Domino IQ Command is set to "Preemptive-AI" (to match what you just created).
- Save the document.
4. Bind Instructions to the Domino IQ Service
- Open any Instructions you want to send to Domino IQ and bind them to the Domino IQ Service. In this example we are binding the AI-Ask Instruction to Domino IQ.
Now, when the AI-Ask instruction is issued, it will be sent to Domino IQ for processing.