Semantic layer · Inference
Intent is the “why now” — as a hypothesis, not a guess.
The inference projection of the AI-facing semantic layer.
Tags say who a member is. Intent reads their live behavior stream to infer why, and why now — high purchase intent, churn risk, coupon-sensitive, advocate. Every signal is stated honestly as a hypothesis to confirm, never as a fact the platform pretends to know.
Member
ada@brand.com
Inferred from the live behavior stream · confirm before acting · read identically by the profile panel, intent tags and the AI agent.
The problem
Behavior is loud, but the “why now” is buried.
A member is browsing, going quiet, or hunting for a deal right now — but most teams only see it weeks later in a report, or bolt on an AI that guesses at raw events with no shared, governed read.
Signals arrive too late
By the time a dashboard shows a member drifting, the moment to act on high intent or catch churn is gone.
Everyone reads it differently
The operator, the segment builder and an AI agent each infer intent their own way — so they disagree about the same member.
Inference gets dressed up as fact
An unlabeled 'this member will churn' invites overreaction. A hypothesis you can confirm keeps the human in control.
The signals
Seven intent signals, inferred from the behavior stream.
Each member carries the intent signals their recent behavior supports — dynamic, refreshed as they act, and always explainable back to what they actually did.
High purchase intent
Behavior points to a member who is close to buying — the moment to make an offer land.
Engaged
Actively opening, clicking and browsing — receptive to the next touch.
Churn risk
Slowing down in a way that historically precedes leaving — a save is still possible.
Coupon-sensitive
Responds to incentives — where a well-timed offer changes the outcome.
Brand advocate
Repeat, high-affinity behavior worth inviting into referral and advocacy.
Newly-acquired
Just joined — the window to shape a habit before it sets.
Disengaging
HoldGoing quiet. Not a target — a signal to hold back and give the member room.
Guardrail: Disengaging is a suppression signal — it tells the platform to hold back and give a member space, never to target them with more outbound.
One core, three surfaces
One inference core — read the same way by people, tags and the agent.
Intent is computed once, in one place. It surfaces three ways so a human, your audiences and the autonomous agent all read identical signals — never three drifting versions of the truth.
A “Behaviors & Intent” panel on the member profile
Open any member and see the intent read alongside their behavior stream — the why behind the signal, in plain language, so a person can confirm it before acting.
Segmentable intent tags
The same signals surface as tags you can build segments and campaigns on — intent flows straight into the governed audience layer next to lifecycle and value.
An agent skill
The autonomous SoClaw agent reads the exact same signals through a governed skill — so the operator, the tags and the AI never work from different intent.
Honest by design
A hypothesis to confirm — not a verdict to obey.
Intent sits in the middle of the layer: below it are the raw behaviors it is inferred from, above it are the governed actions it can inform. Because it is inference, it is weighted accordingly — an agent trusts a certified metric more than a rule-based tag, and a tag more than an intent hypothesis.
Stated as a hypothesis
Every signal reads as 'the behavior suggests…', with the evidence attached — so a person can agree or overrule it.
Confirmed before it acts
Intent informs a decision; it doesn't silently pull the trigger. Money-moving actions still pass through human approval and guardrails.
Weighted by certainty
Metrics are certified, tags are rule-explainable, intent is inferred. Agents read the whole layer and weight each accordingly.
One semantic layer, three projections
State, inference, aggregate — unified by governance.
Intent is one of three governed projections of the same customer graph. Together they are the layer every AI agent, dashboard and API reads the customer through — so they never disagree.
A member's intent read alongside their behavior stream — a hypothesis a person can confirm.
Related reading
Keep exploring the pages most related to this one.
Behavior Capture
One live, unified profile — served through a governed semantic layer, the single interface the AI reads the business through.
Read more CapabilityMembers
One verified, field-encrypted member record fused across every channel into a single behavioral timeline.
Read more CapabilityTags
The state layer of the AI-facing semantic layer: governed, explainable labels for who a member is — lifecycle, value, preference, consent and risk — the certified 'who' every agent, dashboard and API reads and acts on.
Read more CapabilitySegments
Turn One ID profiles and tags into governed audiences: rule-based, query-checked, and ready for activation.
Read more CapabilityCustom Fields
Add your own attributes to the members, stores and products Flash already tracks — a membership number, a preferred store, a child's birthday — and use them everywhere: segments, personalization and import.
Read more CapabilityCustom Objects
Track something loyalty software has never heard of — repair jobs, service appointments, warranty claims — as your own no-code business object: name it, add the fields, collect records, and build audiences from them like any other object.
Read more