Our approach to privacy

SteadMark is built on a simple principle: analyze patterns, not people.

SteadMark reads only the team channels a manager connects, and it does not read channels you have not connected. Direct messages get the same treatment as channels. Patterns merge into aggregates without attribution. Each message is turned into a pattern as it is read, not kept as a message.

What we retain is the pattern itself: the pattern type (for example, “competence uncertainty”), a confidence score, and an anonymous timestamp. The author is represented as a one-way hash, so we can count how many distinct people expressed a pattern without recovering who they are.

We minimize what we store. We keep patterns, not an archive of your messages. Patterns are merged into aggregate totals, and a manager sees the pattern, the source channel, and the date, never message content or sender identity.

We will not sell your data and we will not train shared models on it. You can disconnect Slack at any time from Settings. To request deletion of all stored pattern data, contact us.

What we never do

• We never produce individual employee performance scores, rankings, or risk labels.
• We never share recommendations with HR systems, performance review tools, or compensation platforms.
• We never tell a manager what their direct report said. We tell them what to ask.
• We never use customer data to train shared models.

Compliance roadmap

Regional data residency and customer-managed LLM keys are part of the enterprise architecture. The full security and compliance posture is available on request.

The four commitments

The four commitments that appear across every SteadMark surface are listed in full, with the detail behind each one, on our Privacy commitments page.

For privacy or security questions, contact jim@steadmark.app.