Skip to main content
Profy offers a range of personalization settings that let you adjust how Experts behave according to your needs. These settings affect which model an Expert uses to respond, which tools it can invoke, whether it’s allowed to delegate to other Experts, and other key behaviors.

​
Settings Access

Personalization settings are located on the “Settings” page of the app. Some settings are global (affecting all conversations), while others are per-Expert (affecting only a specific Expert’s behavior).

​
Model Preferences

​
Default Model

You can set a default AI model for conversations. The default model is automatically selected whenever you start a new conversation, saving you from switching manually each time.
Even with a default model set, you can still temporarily switch models in any conversation using the model selector in the input box. A temporary switch only affects the current message and does not change your default setting.

​
Model Characteristics

Different models vary across the following dimensions:

Response Speed

Some models are faster, making them ideal for simple Q&A and real-time interaction

Reasoning Ability

High-end models perform better at complex reasoning, code generation, and long-form analysis

Creative Expression

Certain models are better suited for creative writing, marketing copy, and other creative tasks

Cost Efficiency

Models differ in credit consumption—choosing the right model helps manage costs
The available model list and each model’s specific characteristics are shown in real time within the product. Profy continuously introduces new models for you to choose from.

​
Tool Permissions

Experts come with a rich set of built-in tools (file read/write, code execution, web search, image generation, etc.). You can selectively manage tools for a specific Expert.

​
Plugin Toggles

In an Expert’s settings, you can see its list of available plugins. Some plugins are “user-selectable”—you can enable or disable them as needed:
Controls whether the Expert can build and publish websites. When disabled, the Expert won’t launch website preview and deployment features.
Disabling core tool plugins may affect some of the Expert’s capabilities. For example, disabling the Sites plugin prevents the Expert from helping you build websites. Only disable a plugin when you’re sure you don’t need that capability.

​
Automatic Tool Activation

Most built-in tools don’t require manual management—they activate automatically based on the conversation context:
  • When the Expert needs to search for information, the search tool becomes available automatically
  • When the Expert needs to read or write code, sandbox tools become available automatically
  • When the Expert needs to generate images, the image generation tool becomes available automatically
You only need to manage the “user-selectable” plugins. All other tool activation is driven entirely by the Expert’s judgment.

​
Delegation Authorization

Delegation settings control whether an Expert can delegate subtasks to other Experts. See Multi-Expert Collaborative Delegation for details.
Controls whether your Experts are allowed to initiate delegation. When disabled, no Expert will delegate tasks to other Experts—all work will be completed independently by the current Expert.
When delegation is enabled, you can control which Experts can be delegated to. By default, Experts you’ve used before and publicly available Experts in the marketplace can be delegation targets.

​
Knowledge Base Connections

If you use external knowledge management platforms, you can connect them in Settings to let Experts search and reference your knowledge base content directly during conversations. Currently supported platforms:

Feishu

Connect Feishu documents and knowledge bases—Experts can search and reference your Feishu content

DingTalk

Connect DingTalk documents—Experts can access your DingTalk knowledge spaces

Yuque

Connect Yuque knowledge bases—Experts can search your Yuque documents

IMA

Connect IMA knowledge bases—Experts can query your IMA data
1

Choose a Platform

Go to Settings → Knowledge Base Connections and select the platform you want to connect.
2

Authorize

Follow the prompts to complete the platform’s authorization and provide the necessary credentials.
3

Start Using

Once connected, Experts with the Knowledge Base plugin enabled can search and reference your knowledge base content during conversations.
Your knowledge base credentials are stored encrypted and are used only for conversations between you and your Experts. Other users and Expert creators cannot access your credentials.

​
Settings Persistence

All personalization settings are saved persistently and take effect across sessions and devices:
  • Default model: Automatically applied to all new conversations
  • Plugin toggles: Effective across all conversations with the corresponding Expert
  • Delegation authorization: Global—affects all Experts
  • Knowledge base connections: Global—available to all Experts with the Knowledge Base plugin enabled
If you make a temporary adjustment during a conversation (such as switching models), it won’t change your default settings. Default settings can only be modified through the Settings page.