Skip to main content

AI-Powered Features

AI in practice

Rize uses AI to reduce time-entry cleanup, not to replace your judgment

Rize turns background activity into suggested work entries with titles, descriptions, and tags. The goal is to make time tracking feel automatic while still giving users clear review controls, confidence signals, and privacy boundaries.

AI suggests clients, projects, and tasks from tracked activity
Confidence scores make automation more transparent
Custom instructions help shape titles, descriptions, and matching behavior
Rules and integrations still matter just as much as the model
Rize AI-powered settings
AI settings give teams a practical way to control how much automation they want and where they want that automation to apply.

Video: AI Features in Action​

Auto-Tagging​

Rize's AI automatically tags your time entries with the appropriate Client, Project, and Task. The system now combines multiple signal sources for more accurate matching:

  1. Tag Rules: Your custom tracking rules and keywords provide explicit matching.
  2. Keyword and Embedding Signals: AI analyzes your activity context (app names, window titles, URLs) using both keyword matching and semantic embeddings to infer the best match.
  3. Calendar Events: Meeting titles and attendee information from your connected calendars help the AI assign entries to the right client or project.
  4. External Tasks: Synced tasks from ClickUp, Linear, and Asana provide structured task-level context.

The AI gets smarter over time as it learns from your corrections and work patterns.

Tagging Policy​

Each team can now set a Tagging Policy that controls how auto-tagging behaves for that workspace. The tagging policy lets you define which entity types (clients, projects, tasks) should be auto-tagged and under what conditions.

Tagging Depth​

The Tagging Depth setting controls how aggressively Rize tags your time entries with projects, clients, and tasks. A higher depth means Rize will attempt to tag at more granular levels (e.g., always trying to assign a task, not just a client). A lower depth keeps tagging conservative, only assigning tags when the AI is highly confident.

Configure both settings in Settings > Time Entries.

Confidence score

Each suggestion can carry a confidence score so users understand how certain the system is before they trust or automate around it.

Human review

Rize works best when people review suggestions, especially early in rollout. Fast corrections are part of the training loop.

Task sync context

Synced tasks from ClickUp, Linear, and Asana give the AI richer options to match against for teams with structured project management.

Activity Summaries​

Every time entry receives an AI-generated title and description that summarizes what you were working on. These summaries:

  • Describe the work in natural language (e.g., "Code review on authentication module")
  • Help you quickly understand what happened during each time block
  • Can be customized via Settings > Time Entries > Custom Instructions

Smart Suggestions​

The AI suggestion engine considers multiple signals:

  • Your activity data: App usage, window titles, and URLs
  • Your keyword rules: Custom rules you've created for categorization
  • Your work history: Patterns from past time entries
  • Calendar events: Meeting titles and attendee context from connected calendars
  • Integration data: Tasks from ClickUp, Linear, and Asana

How Suggestions Are Generated​

How AI suggestions are generated
Rize groups work into logical blocks before suggesting titles and tags, which is why good source data and good rules matter.
  1. Rize tracks your in-focus window metadata throughout the day
  2. The AI groups activities into logical work blocks based on context and timing
  3. Each block is analyzed and matched against your Clients, Projects, and Tasks
  4. Suggestions appear in the Suggestion Review Panel on your day calendar for review

Auto-Categorization​

Rize's AI automatically categorizes your apps and websites based on your job title and custom categories. The categorization system:

  • Uses a 75% threshold — if 75% of time in a 15-minute window is spent on focus categories, the entire window counts as focus time
  • Learns from your manual corrections
  • Can be fully customized in Settings > Tracking Rules

Learn more about how categorization works.

Privacy-First Approach​

All AI features are designed with privacy in mind:

  • Rize tracks only metadata (app name, window title, URL) — never the content within windows
  • No screenshots are taken
  • You control what data is tracked in Settings > Privacy
  • You can redact your data at any time
info

The AI runs on Rize's secure servers. Your activity metadata is processed to generate suggestions and summaries, but raw data can be redacted on a schedule if desired.

AI does not change the privacy model

The addition of AI features does not mean Rize starts collecting screenshots or window contents. The product still operates on tracked metadata and user-controlled settings.

Customizing AI Behavior​

Fine-tune how the AI works for you:

  • Custom Instructions: Provide specific guidance for how suggestions are generated (Settings > Time Entries)
  • Suggestion Duration: Set your preferred time entry length
  • Language: Choose the language for AI-generated content
  • Auto-Approve: Enable automatic approval of suggestions for a hands-off workflow

Learn more about Time Entry Suggestions Settings.