Skip to content

Platform · Blocks · AI Enrichment

AI is only as smart
as its context.

Blocks enriches fan profiles with first-party signals and derived attributes inside your environment — so your models and agents reason over governed, structured context instead of raw events and missing fields.

Book a Demo Interview the AI
  • Structured context
  • Derived attributes
  • Consent baked in
  • Your environment

Playable · The interview

Ask the AI anything.
Then change what it knows.

A scripted demo assistant answers three real marketing questions. Flip the context switch between raw logs and governed context — and watch the same model go from shrug to strategy.

What the AI is given

{"uid":"c_8f2","ev":"click","t":…}
{"uid":"sara@…","ev":"scan","t":…}
{"uid":"d_31","ev":"view","t":…}
{"ev":"purchase","seat":"A-114"}
{"uid":"c_8f2","ev":"click","t":…}
… 3,407 more lines …

3,412 lines like this · five “people” who are one fan · no consent fields · no attributes

The interview — demo assistant

Pick a question below. The assistant only knows what the panel on the left gives it — flip the context and ask again.

Scripted illustrative demo — the point is the context, not the model.

Nothing about the model changed between the two answers. The raw logs made it guess; the governed profile let it reason. That gap is what enrichment fills.

How AI enrichment works

Prepare profiles for
intelligent activation

Blocks enriches profiles with derived attributes, predicted scores, and custom features inside your environment — so every downstream model and agent starts from a clean, governed, structured signal.

  1. 01

    Structured context, not raw events

    Profiles are enriched with first-party signals and derived attributes before they leave your environment — models receive governed, structured input instead of raw logs with missing fields.

  2. 02

    Privacy-preserving enrichment

    Enrichment runs inside your environment on data you own and control. No vendor sees the underlying fan graph, and consent rules are enforced at every enrichment step.

  3. 03

    Context stays in sync

    Enriched attributes land on the same profile that drives segmentation and activation — so AI context and activation context never diverge.

  4. 04

    Custom and standard attributes

    Apply off-the-shelf derived attributes — predicted churn, propensity to spend, lifecycle stage — or build your own through the API, so context fits your models and strategies.

The transformation

What enrichment actually adds

Three layers turn a noisy event log into something an AI can reason over — each one computed inside your environment, on data you own.

Layer 01 · Identity

Five records become one person

Events from the cookie, the inbox, the app, and the gate are deduplicated onto a single fan — so the model never counts one supporter as five strangers.

5 uids → Fan #48211

Layer 02 · Derivation

Behavior becomes attributes

Raw scans and sessions are distilled into derived attributes — matchday affinity, lapse risk, propensity to spend, lifecycle stage — that a model can use directly.

312 events → affinity: high · lapse risk: low

Layer 03 · Governance

Consent travels with context

Consent and residency flags are stamped onto the profile itself, so every answer an agent gives is already inside the rules — compliance is an input, not a review meeting.

consent: email ✓ push ✓ ads ✗

Why it matters

Raw signals in,
AI-ready intelligence out

Raw events are noisy and fragmented. Blocks transforms them into clean, structured attributes — predicted churn, propensity to spend, lifecycle stage, custom features — so your AI systems focus on intelligence, not data cleaning. All inside your environment.

See how agents use enriched profiles
  • Accelerate AI model performance by feeding clean, structured input instead of raw event streams.
  • Reduce model training complexity with governed, first-party features instead of inferring intent from fragmented logs.
  • Stay compliant by enriching with consented data in your environment, not a third-party platform.
  • Keep profiles fresh by updating enriched attributes in real time as new signals arrive.

Give your AI context worth reasoning over

See Blocks turn raw fan events into governed, enriched profiles inside your environment — and watch what your models can do with them. Book a walkthrough with our team.

Book a Demo