All case studies
LiveRegulated market

Guardian Compliance

From paper records to 950 live sites.

950

sites live as of

35%

of the sites in the market

Problem
Paper or legacy records, no cross-venue trail
Move
One customer record across venues
Role
End-to-end product managerMerkur account owner
Timeline
Aug 2024 to 19 Oct 2026

Venue by venue to one record

The question: what do staff on the floor need to notice someone in trouble, and what does a compliance team need to show they acted?

Customer record
Paper formsInteraction logLegacy systemKYC checksVenue by venueAcross venuesStaff miss itMachine alertsNo audit trailAudit trail

Click through it

Hit play, then click the pulsing dot.

guardian, sanitised demo, fictional dataPlaceholder

Guardian Compliance click-through

Sanitised wireframes, fictional data

add: the real click-through goes here

Three decisions

Play with the first. Tap boxes on the others.

Prompt people, don’t decide for them

Operators set the lines. Staff get the prompt. Configurable rules, not AI.

Net loss£0
Time on game0 min
Cash in£0
  1. Alert
  2. Staff prompted
  3. Interaction logged

Fictional data, not the Guardian UI. Dark lines are the operator’s thresholds.

add: why rules were right hereadd: how loose settings are guarded

Through the trade body

The route to market was our MD’s call. I owned the Merkur account.

MD’s pitch distribution I owned the account 1 customer, not ~500 RBW Bactatrade body Merkur members’ sites, ~700

Swipe the diagram →

Tap a box.

add: my part in the rolloutadd: what changed because of Merkuradd: how many of ~700 are live

Build to rollout, to scale

Drawn to scale: 26 months from first build to 950 live sites.

Build beginsAug 2024PilotsOct 2025Wider rolloutJun 2026950 sites live19 Oct 2026

Swipe the diagram →

add: sites live at each stage

One operator’s inspection

Anonymised.

  1. Special measures
  2. Rolled out Guardian
  3. Cleared

Tap a step.

sensitive: check wording against the Commission’s guidance and RBW, or cut

Next

add: what I’d change, and what’s next for the product

Next case study

RBW Atlas: one system of record, and an agent that knows when to ask a person

Read it