Automatic time tracking

Time tracking that
fills itself in

See how many hours every person spends in every app - rebuilt automatically from the activity logs your tools already keep. No timers to start, no timesheets to chase, no screenshots.

  • No timers
  • No screenshots
  • Metadata only
collaborationsonar / people / maya-chen
Maya Chen Eng Lead, Payments · Tuesday
8h tracked today
  • Code editor 3h 45m
  • Meetings 1h 50m
  • Slack 55m
  • Jira 35m
  • Browser 30m
  • Outlook 25m

Rebuilt from the activity logs of the tools you already use

SlackJiraGitHubConfluence OutlookGoogle WorkspaceMicrosoft TeamsZoomLinearFigmaNotionAsana
Why it's different

Timesheets are fiction. Your tools already know the truth.

Classic time tracking asks people to report their own hours. We read the record that already exists.

Classic time trackers
  • Timers people forget to start - and forget to stop
  • Friday-afternoon timesheets reconstructed from memory
  • Screenshots and keystroke logging that erode trust
  • Hours you can't break down by what the work actually was
CollaborationSonar
  • Time is rebuilt automatically from activity your tools already log
  • Accurate to the session, not to someone's best guess
  • Metadata only - no screenshots, no keystrokes, no content
  • Every hour attributed to the app it was spent in
How it works

From audit logs to hours, in three steps

Connect once, backfill history, and see the first weeks of data in days.

  1. 01

    Connect your tools

    Read-only OAuth to Slack, Jira, GitHub, Outlook, Google Workspace and the rest. No timers to roll out, nothing for people to fill in.

  2. 02

    We rebuild the working day

    Activity timestamps are unified per person across tools and stitched into sessions - so a morning of commits reads as a morning in the editor.

  3. 03

    See where time goes

    Hours per person, per team and per app. Focus time, meeting load and context switching - trended week over week.

Team view

Every hour of the week, broken down by app

Who is heads-down in code, who lives in meetings, and who spends the week answering Slack - side by side, for the whole team.

  • Time per applicationPer person, per team, per week.
  • Focus vs meeting loadTrended week over week, so you see when it tips.
  • Unified identitySame person across Slack, GitHub, Jira - one timeline, not five.
collaborationsonar / teams / growth-and-platform
Time by app Last 7 days
Tracked 236h 6 people
Focus apps 39% +4 pts vs last wk
Meetings 17% +2 pts vs last wk
Hours per app, last week
  • Code editor
  • Slack
  • Browser
  • Jira
  • Meetings
  • Figma
  • Outlook
  • Other
View as table
Person Code editorSlackBrowserJiraMeetingsFigmaOutlookOther Total
Maya Chen 14.5h5h4h3.5h9h0h2h1.5h 39.5h
Jakub N. 24h4h6.5h1.5h3h0h1h1h 41h
Aisha O. 0h7.5h7h8h12.5h2.5h4h1.5h 43h
Lena W. 0h5h5.5h2h6h19.5h1.5h1h 40.5h
Tom P. 11h3h5h1h4h1.5h1h2.5h 29h
Ravi K. 19h6.5h8h2h5h0h1.5h1h 43h
What you can answer

Questions timesheets never could

When hours come from real activity, you can finally ask what they were spent on.

Is the team drowning in meetings?

Meeting load per person and per team, next to the focus time it replaced.

How much did the contractor actually work?

Contractor and agency hours without chasing timesheets - evidence, not self-reported totals.

Where does engineering time go?

Editor vs Slack vs Outlook vs Jira vs meetings - the real split behind "we're too busy to ship".

Who is context-switching all day?

Fragmented days show up as dozens of short sessions. Spot them before they burn people out.

Is the new hire ramping up?

Watch time shift from docs and Slack into real work over the first weeks.

What does remote actually look like?

Working patterns across time zones, without anyone reporting their hours.

Privacy & security

Time tracking people don't have to hate

No screenshots, no keystroke logging, no reading messages. Built to pass your security review.

Metadata only

Timestamps and action types. Never message bodies, code, documents or screens.

Read-only access

OAuth scopes only request read. We never write to your tools.

You stay the controller

Built on data your organisation already lawfully processes.

Flexible hosting

Dedicated environment on SOC 2-compliant infra, or on your own.

SSO / SCIM

Okta, Entra ID, Google Workspace.

Audit log of audits

Every query and export is logged.

FAQ

How the time gets counted

The questions we hear most. See all FAQs →

How do you know how much time someone spent in an app?

From activity metadata - timestamps, user IDs and action types - that your tools already record. We unify each person's identity across tools and stitch nearby events into sessions, attributed to the app they happened in.

Do people have to start timers or fill in timesheets?

No. Nothing changes in anyone's workflow. Time is reconstructed from the tools you connect; there is nothing to click and nothing to forget.

Do you take screenshots or log keystrokes?

No. We only read metadata: when something happened, who did it and what kind of action it was. We never ingest message bodies, file contents, source code, document text or screen captures.

What about time spent outside the connected tools?

It isn't counted. CollaborationSonar only sees activity in the tools you connect, so offline work - whiteboarding, reading, thinking - shows up as a gap, not as invented hours.

Where does my data live?

By default, in a dedicated environment on our own SOC 2-compliant infrastructure. If you'd rather keep everything in-house, we can deploy your environment on your own infrastructure instead.

From the blog

Notes from the field

All posts

Why we're building CollaborationSonar

A short note on why operational visibility matters more than people management dashboards.

Stop chasing timesheets

Connect your tools and see where your team's week actually goes - without asking anyone to track a single minute.