SocialHub.AIFlash

Loyalty & Members · Members

One member record across every channel — not the same customer counted twice.

A buyer scans in store, logs into Shopify, opens an email, sends a referral — and most stacks see four strangers. Flash fuses them into one verified, encrypted member, with every touch stitched into a single timeline.

Identity fusion · one verified record
In-store QRShopify SSOEmbed SSOEmail
AES-256-GCMBlind indexTakeover guard

The problem

Fragmented identity is invisible churn.

When the same person shows up as separate rows per channel, every report is wrong, every send is duplicated, and your best customer looks average across four half-empty profiles.

One customer, six fragments

Store, QR, Shopify, receipts, email and marketplace each mint their own record. Nobody is the real customer.

Spend you can't see

Purchases on one channel never line up with engagement on another — so lifetime value is a guess, not a number.

PII that's a liability

Customer email and phone sitting in plain columns is a breach waiting to happen — and a compliance question you can't answer.

How it works

Fuse the identity. Encrypt the record. Stitch the timeline.

Identity fusion happens at the global users layer, so one email is one person — while program data stays isolated per team. Every PII field is encrypted, yet still searchable. Every touch merges into one chronology.

Field-level encryption

Searchable, but never stored in the clear.

Emailblind index
Email local partblind index
Full name3-char prefix index
Phone

Each PII field is encrypted with AES-256-GCM. Exact-match search runs over an HMAC-SHA256 blind index from a dedicated sub-key — so members stay findable by email or name without the plaintext ever being indexed.

Unified timeline

Ten-plus event types, one chronology.

Enrollmentconsumer-enroll
QR scancaptures
Points earnedledger
Email openedtracking · device + proxy flag
ReceiptOCR pipeline
Purchaseflash_orders
Coupon redeemedwallet
Portal visit / clickpage_visits · block events
Referral convertedreferral_invitations

Enrollment, scans, points, email opens (with device and mail-proxy flags), receipts, purchases, coupons, portal visits and clicks, referrals — merged from their real sources, newest-first. Purchase and receipt stay deliberately separate rows.

Conservative fusion

enrollConsumerByEmail merges only at the global users layer — idempotent on (program, email), race-safe, with welcome points and consent recorded once.

Takeover guard

Internal staff accounts are refused at enrollment, so a consumer flow can never quietly take over an internal identity.

Behavior tracking, attributed

Page visits, block-level portal clicks, and email opens — with device and mail-proxy detection — all attribute back to the real member.

Why it's different

A verified member graph, not a stack of synced lists.

The enterprise way is to assemble identity, analytics, and decisioning across separate products. Flash unifies the member as one verified identity across store, QR, Shopify, receipts and marketplace — by design.

Typical approach

CDP + engagement stack

Identity resolution and analytics live in separate products you integrate and reconcile.

Flash, by design

One verified member object across every channel, encrypted field-by-field at the source.

Typical approach

ESP list sync

Each channel keeps its own list; the same person appears many times.

Flash, by design

Fusion at the global users layer means one email is one member — conservatively, with a takeover guard.

Typical approach

Plaintext PII columns

Customer data sits in the clear, searchable but exposed.

Flash, by design

AES-256-GCM per field, with HMAC blind indexes so search works without ever indexing plaintext.

Insight & innovation

One verified record makes the analytics worth trusting.

Because every touch resolves to one member, the per-member insight bundle runs on real, complete data — and degrades safely rather than fabricating when a signal is thin.

RFM, CLV & health score

Per-member recency/frequency/monetary, projected CLV, and a 0–100 health score — each with a confidence level, never a false-precision number.

Lifecycle stages

A decision-tree stage — new, growing, mature, at-risk, dormant, churned — so the next action is grounded in where the member actually is.

Relationship graph

A daily pre-computed snapshot of the referral network with a zero-dependency force-directed layout — see who recruited whom.

The insight bundle runs each engine independently — one failing engine degrades to a low-confidence default rather than failing the whole profile. Honest by construction.

What changes for the business

You stop paying to reach the same person twice — and start treating your best customers like your best customers.

AES-256-GCM

field-level PII encryption with 3 blind indexes

10+ events

merged into one member timeline

One identity

fused across store, QR, Shopify, receipts & marketplace

See one member, instead of six fragments.

Walk a real cross-channel identity through fusion, encryption, and the unified timeline — live.