Compass helps you localize the decision. Centaur holds the move against a precommitted threshold, so when the real signal arrives you can compare expected to actual instead of relying on memory, mood, or hindsight. Founder access begins after Compass, while the advisor workspace is still rolling out through a reviewed waitlist.
Active episode
In progressCurrent intervention
Double down on the PLG motion — delay outbound
Testing hypothesis
If self-serve is the constraint, PLG removes it within 3 weeks
Success signal
Week 3: 4+ new orgs activate without sales touch
Review date
May 14, 2026
How it works
Import directly from your Decision Brief link. The decision frame, readiness, and likely explanation are already loaded.
Define the result that would strengthen or weaken the move. Set the date before the evidence arrives.
Compare expected to actual, record the learning, and let the next decision start from sharper context.
How you arrive
The readiness verdict changes the job. Sometimes Centaur tracks a committed move. Sometimes it tracks the test you still need before the real decision is ready.
Verdict: Ready
You have enough to decide. Centaur records the chosen frame, the primary hypothesis, and the threshold date.
Typical: days to two weeks
Verdict: Not yet
Run the discriminating test first. Centaur tracks what you ran, what came back, and whether the evidence was enough to upgrade the decision.
Typical: one to three weeks
What Centaur tracks
Not a project manager. Not a CRM. Centaur keeps the expected signal, the actual result, and the learning in the same place so the case stays legible over time.
The decision you committed to
The exact frame you chose, linked back to the Decision Brief it came from.
The hypothesis you are testing
The explanation you are betting on, with the evidence that would strengthen or weaken it.
The threshold signal
The result that tells you whether the move worked, defined before action so it cannot drift later.
The learning
What is more true now than before. This keeps the next episode from starting blind.
Copilot
Copilot sits on top of the episode, not beside it. Use it to stress-test a hypothesis, interpret a borderline threshold result, or identify what evidence is still missing — grounded in the current case, not generic AI advice.
For advisors
Episode history makes follow-up smarter. What the client committed to, what the signal was, and what actually happened all travel into the next session, so your advice compounds instead of restarting.
Advisor access is rolling out through a reviewed waitlist — we're notifying advisors as the workspace opens.
Start with Compass to localize the decision. Use Centaur once the move needs follow-through, thresholds, and learning over time.