Introduction

Constructor Model is a solution that allows you to create and manage knowledge models. These models organize and manage uploaded source data so that artificial intelligence (AI) can understand and use it when interacting with users, for example, answering their questions, generating text, and so on.

In Constructor Platform, knowledge models are powerful tools designed to enhance learning, teaching, and research. They provide learners, educators, and researchers with quick and easy access to subject-specific information. By using knowledge models, users can explore topics more effectively, deepen their understanding, and simplify their studies or research.

Two types of knowledge models are available:

• Knowledge Assistant — standard knowledge models that make it easy to turn your data into interactive AI tools such as website assistants, chatbots, or tools for research and education
• Conference Assistant — custom knowledge models designed to support scientists attending conferences by providing detailed event information, speaker bios, and personalized agendas with recommended lectures and workshops.

How to use knowledge models

Create your knowledge model and upload your own files containing source data for the model to use in generating answers. The model will analyze the contents of these files based on keywords, topics, and their relationships.

Enter your question in the chat input field. For more accurate results, provide sufficient context and use plain language. The model will reply based on the data from the files, and each response will reference the files used to generate it.

You can use knowledge models in various scenarios, such as:

• Organizing your gathered research notes for easy access to necessary information and facts.
• Building a library of educational resources and scientific papers to share with your learners for theoretical study.
• Generating learning content for your courses or other educational materials from the source files you upload.
• Generating scientific hypotheses based on your data.
• Embedding them as AI assistants on your websites.

Alternatively, you can choose to chat directly with the large language model (LLM) engine selected for your model (such as Google Gemini or Anthropic Claude). In this mode, no uploaded files are used.

Data access levels in knowledge models

You can create various knowledge models based on different sets of data and enable or restrict public access to them. Constructor Model ensures the security of your data by providing flexible configuration of user access permission levels.

How to get started with knowledge models

For quick start information on using knowledge models, see the following:

• Registering in the platform with your email
• Registering in the platform with external accounts
• Getting started with knowledge models

For the information on how to use knowledge models, see Managing knowledge models.