brainis
Product Updatesproduct-updatescortexautopilot

Introducing Cortex Autopilot

Scheduled AI analysis with guardrails: pick the plays, set the cadence, and get results as digests instead of dashboards.

B
Brainis Team
August 15, 20261 min read · 261 words

Cortex Autopilot runs the analysis you would run every Monday, without anyone remembering to run it.

What it does

An Autopilot rule pairs a play with a schedule. "Run the pipeline-risk diagnostic every Monday." "Check cash-flow projection on the 1st." Each pass executes, records its findings, and delivers them through your digest rather than into a dashboard you would need to remember to open.

The guardrails

Autopilot runs under the same authority contracts as every other AI surface in Brainis:

What you'll learn
  • Read-only by default. A pass analyzes and reports. It changes nothing unless its autonomy level allows it.
  • Approvals for actions. If a pass concludes something should change, the proposal lands in your approvals queue unless you have explicitly delegated that verb.
  • Rate and budget caps. Contracts can limit executions per hour and monetary impact, and a paused rule costs nothing.
  • Full audit. Every pass appears in the Cortex Fleet with step-by-step replay.

What it costs

About 6 credits per pass. A weekly rule is roughly 24 credits a month. Scheduled Autopilot runs are available on the Business plan; the same plays can be run manually on any plan.

Getting started

Open Cortex, choose a play you already run manually, and add a schedule. Start with one. The most common mistake is scheduling six analyses in a week and reading none of them.

Tip: Review your Autopilot rules quarterly. The failure mode of scheduled analysis is not wrong answers; it is nobody reading right ones.

Read the full guide in the Autopilot documentation.

product-updatescortexautopilot
B
Brainis Team

Sharing insights on business operations, AI, and modern team management.

Run your company on Brainis

All 11 Operating Systems on every plan, from $29 a month. No per-seat pricing — you pay for AI capacity, not headcount.

See pricing