Data Linker - User Documentation
  • Data Linker
    • Introduction to Datalinker.AI
    • Quick-start guide on using the application
      • Signup/Registration
      • Logging In
      • Forgot Password
      • Create an Organisation
      • Create a Database, Chat Category & Assign users
      • Chat with Chat Category
      • Invite users
    • Signup & Registration
      • Manual Sign-Up
      • Invitation Sign-Up
    • Create an Organisation
      • Fill in your organisation details
    • Organization Management and User Roles
      • Organization Management
      • Inviting Users to an Organisation
      • Managing Users and Roles
      • Managing Billing & Token Usage
      • Understanding Token Usage
      • Input & Output Token Definition
      • When Tokens are Depleted
      • Topping Up Tokens
    • General setup steps
    • Create a Database
      • Creating a New Database
      • Managing Databases
    • Chat Categories
      • Adding a New Chat Category
      • Managing Chat Categories
      • Interact (chat) with the Chat Category
    • “Threads” within Chat Categories
      • What is a “Thread”
      • How should I utilise the thread function?
    • Chat Category Collaboration
    • AI Models (LLMs)
      • Switching AI Models
    • Additional information: Instruction Fields
      • Purpose of the Instruction Field
      • Shaping AI Behaviours
    • Disclaimer
  • DATALINKER - TERMS, PRIVACY POLICY
    • PRIVACY POLICY
    • TERMS OF USE
    • TERMS OF SERVICE
Powered by GitBook
On this page
  1. Data Linker
  2. Additional information: Instruction Fields

Shaping AI Behaviours

  • By entering instructions in the field, users can instruct the AI to assume a particular role or perspective.

  • For example, users can specify that the AI should act as a chatbot, an AI assistant, or a subject matter expert in a specific domain, additionally users can set rules such as “as an AI company chatbot you will only give answers relating to” Company Name” and the embedded information here within” This helps align the AI's responses with the user's expectations and the intended purpose of the conversation.

  • The instruction field can also include additional details about the document being embedded. Users can provide relevant information such as the document's source, purpose, or specific sections they want the AI to focus on.

  • These details help the AI understand the user's needs and provide more accurate and tailored responses.

IMPORTANT! Understand that while the instruction field helps guide the AI's behaviour, the AI's responses are generated based on the training data and patterns it has learned from. The AI may not always provide perfect or complete answers, and it is recommended to review and validate the information obtained from the embedded documents when necessary.

PreviousPurpose of the Instruction FieldNextDisclaimer

Last updated 1 year ago