AI Features in Userback

Userback uses AI in three places today: Identifying the type (Bug, Task, Idea, General), categorizing it highlighting the sentiment, determining if a reply is required, surfacing patterns across a lot of feedback at once, and helping you write replies faster.

You have full control over AI features in your workspace and can enable or disable them at any time. The toggle exists in your Workspace settings.

From the left nav click your profile on the bottom left > Workspace Settings > Navigate to AI Features > Toggle AI features on or off.

AI Assist automatically classifies incoming feedback as a Bug, Task, Idea, or General item as soon as it's submitted, instead of leaving every new item as "unsorted" until someone manually tags it. It's confirmed on by default.

The toggle lives on Feedback > Settings > Widget > Design > Feedback Types. When it's on, the manual type-configuration section below it (checkboxes and labels for each feedback type, plus a default-value setting) is grayed out, since AI Assist is handling classification instead.
​
Turn AI Assist off to configure those types and defaults manually.

AI Assist will also categorize feedback (Tag) based on existing categories that you have used previously on feedback. The feature is labeled Auto Categorization ("Use AI to automatically categorize new feedback"), which also exists on Feedback > Settings > Automation. Enable this toggle to have all feedback automatically categorized for better searchability.

AI Themes groups similar feedback into themes automatically, using AI, so you can see patterns across dozens or hundreds of items without reading each one. Open it from the AI Themes button on the Feedback board. See Understand AI Themes for the full walkthrough, including theme detail views, evidence, and archiving.

When you're writing a comment or internal note on a feedback item, a sparkle icon in the comment toolbar opens an AI assist menu with three options: Fix spelling & grammar, Improve writing, and Simplify text. Each rewrites your current draft in place. This is covered as part of the comment workflow in Comments and Notifications, not repeated here.

A related setting, AI Tone, exists under My Profile ("Personalize how AI drafts replies for your workflow"). This affects how the reply-drafting assist rewrites your text, and is a per user setting.

Userback uses leading AI model providers to power our AI features, currently including OpenAI and Anthropic. Providers are selected for their performance in natural language processing and enterprise-grade security standards.

All AI models run within Userback's dedicated cloud infrastructure on AWS. This architecture means that processing happens in our controlled environment rather than on the AI provider's public consumer platforms. We maintain strict security controls over data in transit and at rest, ensuring your data is subject to the same security standards as the rest of the Userback platform.

Read more about our Security and Privacy Policies.

Your data is never used to train AI models. When you use AI features in Userback, your feedback data is processed solely to generate insights for your workspace.

Our AI providers do not use your data to train or improve their models, and your data is not shared with other Userback customers or third parties. Data processing only occurs when you actively use AI features.

Userback MCP (Model Context Protocol) lets you connect your AI tools and code editors directly to Userback, so you can work with feedback, bugs, and project context without leaving your development environment.

Configured clients include Claude Desktop, Claude Code, Cursor, ChatGPT, VS Code, and Windsurf.

With Userback MCP you can:

For setup instructions, see the Userback MCP documentation.