Signals no single view can see alone.
One layer watches health, another risk, another your calendar. Insights reads them all at once — and catches the things that only show up in the combination.
Cross-layer synthesis, after every email
After the other layers finish their work on an email, Insights reads across all of them together. A rising health score is nice; late-stage language is normal; a CFO added to the thread is routine. All three at once, with a fiscal deadline days away, mean one thing: a close window. It's the combination that's predictive.
- Runs after every other layer commits — it reads their conclusions
- Fires only on multi-layer combinations, which is what gives it signal
- Lands in your brief with the exact signals that produced it
The patterns it watches for
Close windows, pre-churn, aligned deal risk, re-engagement spikes, fundamental behavioral shifts when someone changes role or team, and seasonal rhythms — like a contact who always goes quiet in Q4, so that silence is correctly read as normal, not alarming.
- Close-window, pre-churn, deal-risk, re-engagement, behavioral-shift
- Seasonal patterns prevent false alarms on predictable quiet spells
- Each becomes a permanent, high-confidence memory about that contact
It keeps score on itself
Every prediction is tracked against what actually happened, and your one-tap feedback on each decision feeds the same loop. Patterns that prove right earlier earn lower thresholds; ones that misfire get raised. No model retraining — just honest, personal calibration that makes the warnings sharper the longer you use it.
- Every prediction recorded and later resolved against reality
- Your ✓/✕ on decisions trains its calibration
- Thresholds self-adjust per pattern — sharper over time
Good to know.
How is this different from relationship health?
Health watches one dimension per contact. Insights reads health, risk, calendar and memory together — it catches things like a close window or aligned deal risk that no single layer can see on its own.
Does it get better over time?
Yes, specifically for your world. It tracks whether each prediction came true and folds in your feedback, then adjusts its own thresholds — no retraining, just statistical calibration to how your relationships actually behave.
Where do its signals show up?
In your morning brief, each with the underlying signals that triggered it, so you can act in one click — and see, later, whether it was right.
See it on your email.
Connect Gmail and it reads your last 90 days. Your first brief shows you what it already found.
Start free — connect GmailNo credit card · Personal email permanently excluded · Export or delete everything anytime