Cloudflare Pages prototype

Runner turns Melissa/Unison data quality tools into an augmented decision workflow.

This static prototype shows the missing AI authority layer: profile evidence becomes a recommended, reviewable, auditable package that existing Unison services can execute after steward approval.

Unison ADQ Package Authority

Profile metrics are converted into a task plan with confidence, field mappings, review gates, task dependencies, change codes, and a package hash.

Open ADQ demo

ADQ Execution Interface

Runner EDQ acts as the interface while microservices check emails, verify phones, geocode addresses, cluster duplicates, and produce golden records.

Open execution demo

Runner + Melissa Strategy Map

The merger story is mapped to practical assets: Runner supplies the AI product layer while Melissa/Unison supplies proven verification, matching, and survivorship services.

Open strategy map

Wirecontracts

The architecture, steward cockpit, and primitive/component contract for building the ADQ layer with modern React product primitives.

Open wirecontracts

What this demonstrates

Augmented search Find the problems worth acting on from profile evidence.
Package assembly Bundle profiling, verification, matching, and survivorship into one plan.
Steward control Keep humans in the approval loop before data mutation.
Local learning Improve from accepted, edited, and rejected recommendations.

Demo path

  1. Open the ADQ package demo and inspect how evidence becomes executable metadata.
  2. Open the execution demo and inspect how service calls produce golden records.
  3. Open the strategy map and connect each Runner layer to a Melissa/Unison capability.
  4. Use the generated JSON artifacts as the contract for the next real Unison integration.