Analytics and catch-up

Three different questions, three different views, and it is worth knowing which one you are asking.

Catch-up, “what did I miss?”

A digest since a point in time: decisions approved, incidents opened, approvals resolved, notable activity. Built for the person coming back from a week off, or joining a project mid-flight.

Console → Dashboard, the catch-up panel showing what changed since the last visit

screenshot slot, save as public/screens/…png and set src

Catch-up is scoped to what you can see. A viewer's digest and an admin's digest are different documents.

Analytics, “how are we doing?”

Volumes and rates over time: actions by effect and risk, approval latency, incident counts, ingestion throughput, decisions approved and rejected.

The number worth watching first is approval latency. Rising latency means a rule is routed to people who are not looking, which ends with somebody disabling the rule rather than fixing the routing.

Network insight, “what spans workspaces?”

Cross-workspace patterns: which teams' work depends on which others', where the same decision keeps being re-made in different corners, where two groups are governed inconsistently.

The report bundle

One command, everything an auditor usually asks for in the first meeting: policies in force with their version, decisions with approvers, incidents with resolutions, and the audit chain's verification state.

bash

memnox report                   # the evidence bundle, markdown or JSON
memnox compliance controls      # every mapped control, its status and evidence
memnox compliance summary       # one line per framework, for a weekly read
memnox insights                 # what is being attempted, blocked and approved

Those are the runtime's own, built from the audit trail on your machines. The console produces the same bundle for the whole organization on Settings → Reports.

See For compliance and legal.

What analytics deliberately does not do

No forecasting. Memnox reports what happened. A projection is a model's opinion, and there is no model in this path.
No individual productivity metrics. The graph knows who did what. Turning that into a ranking of people is a use this product is not built for, and the data would be wrong for it anyway, ingestion sees the tools that are connected, not the work that happened elsewhere.
No scoring of teams. The same reasoning.

Usage and quota

Settings → Billing shows what has been consumed against each of the plan's limits this period, and resets with the billing period.

The number to watch is governed actions: it is the one that grows with adoption rather than with configuration. A request that would exceed a limit is refused with a clear reason rather than being silently truncated, because a half-read history that looks complete is worse than being told no.

For leadership

The weekly read, and what to ignore.

Activity and audit

The underlying record everything here is computed from.