Docs
Not watched. Coached.
Most tools in this market install a desktop agent and watch the screen. InRoots reads one line a day in chat and explains a score. The differences follow from that.
Input
- Agent-based tools: desktop agent, screenshots, app and URL logs, sometimes keystrokes, sometimes live screen
- InRoots: a one-line check-in in chat or spoken to the assistant; presence confirmed with a consented ping
What is measured
- Agent-based tools: activity proxies (active time, app categories) classified by a black box
- InRoots: outcomes (tasks shipped, blockers cleared, on-call coverage, objectives) cross-checked against Jira, GitHub and GitLab, weighted in the open
Who sees what
- Agent-based tools: managers see everything; employees see little or nothing
- InRoots: the employee sees their own scorecard first; the lead and the team see the same number; manager-only views are consent-gated and explained
AI
- Agent-based tools: AI classifies apps as productive or not, usually through a public API
- InRoots: 16 providers or a local model; the score is deterministic math and the AI explains and coaches
Where the data lives
- Agent-based tools: vendor cloud, sometimes with an on-premise option at the top tier
- InRoots: self-hosted in one container or hosted by us, same features either way, no telemetry
The result
Teams accept what is fair and visible. That is the whole difference, and it is why InRoots calls itself a productivity app, not a monitoring tool.