Knowledge

Creating Custom Models in Collective

Build AI models tailored to your organisation’s needs — learn how to train, configure, and deploy custom models within Collective.

Overview:

Custom models in Collective are specialised AI language models tailored to meet the specific needs of an individual or organisation. These models can be fine-tuned with unique datasets and system prompts.

Use Cases for Custom Models:

  • Industry-Specific Applications: Healthcare, banking, etc.
  • Content Generation & Branding: Articles, social media posts, etc.
  • Technical Writing & Knowledge Management: User manuals, technical documents, etc.
  • Sales & Marketing: Email campaigns, sales scripts.
  • Financial Analysis & Reporting: Financial reports, market trend analysis.
  • HR & Recruitment: Job descriptions, resume screening.
  • Legal Assistance: Legal document drafting, case summaries.

How to Create a Custom Model

  • Navigate to the Workspace > Models tab.
  • Click “Create a Model.”
  • Upload an image/icon for your custom model.
  • Choose a base model and define the system prompt.
  • Add knowledge (e.g., documents) to your model.

Testing and Refining Your Custom Model

Test the model by refining its system prompt and base model to improve the responses.

‍

Related articles

Using the Document Vault in Collective
Creating Custom Models in Collective
Prompt Guidance for Collective