Bounded scope. Controlled capture. Team-level reporting.

1. Purpose-led

Every Study starts with an agreed business question, team or workflow, reporting audience, time window, and evidence boundary. Teho is not used for open-ended monitoring or "let's see what turns up" investigations.

2. Minimum capture

Capture is governed by the agreed Study boundary. Out-of-scope and non-work activity is excluded from the evidence base, and deployment-specific controls define what is technically blocked before capture. Clients can define additional exclusions upfront, including applications, keywords, time periods, and sensitive windows through privacy-guard rules.

3. Optional evidence by agreement

Some deployments may use richer evidence, including screenshot, semantic, or LLM-enabled classification paths, where that evidence is needed to answer the business question and is explicitly approved by the client. Those settings are configured before rollout, named in deployment materials where relevant, and can be disabled where they are not required.

4. Out of reach by design

The agent is not designed to capture keystroke transcripts, credentials, microphone audio, camera video, open-ended device history, or a continuous visual memory of everything a participant sees or does.

5. Team-level reporting

Study outputs are designed for team or workflow-level operating insight. Only reportable in-scope patterns that pass privacy, coverage, and confidence checks become leadership evidence. Minimum cohort thresholds suppress slices that are too small. Teho does not produce individual productivity scores, rankings, leaderboards, pace metrics, or supervisor-facing activity feeds.

Final readouts should be interpreted with their Coverage & Confidence notes: some real activity may be excluded, below threshold, unclassified, low-confidence, or outside the observation window.

6. Client-controlled

Deployment scope, capture settings, consent posture, visible indicators, privacy-guard rules, reporting audience, retention, deletion, subprocessors, and transfer paths are agreed before deployment and held stable unless a formal change is approved. Those controls are client-specific and should not be read as identical across every Study.

7. Retention and deletion

Retention is set in the deployment-specific control pack and DPA. Screenshots, if enabled, can be purged after classification or retained for the agreed window. Observation data is deleted after the agreed retention period, with destruction evidence available on request.

8. Contact

For Study privacy questions: info@teho.ai. For engagement-specific questions, the client sponsor receives the named Teho lead and privacy contact during scoping.