Platform

The platform architecture.

The seven-stage lifecycle:Every account moves through seven automated stages that resolve raw web context into verified AI truth.

Account command centerthe view an account manager works from

1 of 11 · Command centre

Seven stages. One operating graph.

ACE ingests business signals, builds context, reasons over that context, coordinates execution, and learns from the outcome. Its growth modules operate across that same graph.

  1. 01

    Ingest

    Bring every relevant signal into the operating context.

    CRMCalls and conversationsWeb activityPaid mediaSearch demandContent performanceIntent signalsConversion data
    HubSpot · GA4 · Google Ads · Meta · TikTok · Apollo · RB2B
  2. 02

    Enrich & Normalize

    Validate, standardize and complete the data before ACE reasons over it.

    Contacts and companiesCampaign taxonomyChannel taxonomyNaming conventionsAudience structureAttribution fieldsFirmographics
    Clay · Waterfalls · ACE normalization rules
  3. 03

    Build Context

    Resolve structured and unstructured data into the operating graph: people, companies, campaigns, channels, interactions, intent, performance and outcomes.

    Knowledge GraphRAGHistorical stateBusiness rulesRelationships across systems

    This is also where the Campaign Blueprint Engine defines the operating structure ACE expects:

    Campaign naming conventionsChannel taxonomyAudience structureBudget groupingAccount configuration
  4. 04

    Evaluate & Reason

    Continuously evaluate the graph to determine what is healthy, what changed, why it matters, and what should happen next.

    Tracking Readiness Engine
    GA4 eventsGTM tagsGoogle Ads conversionsMeta Pixel + CAPIAttribution integrity
    Pre-Launch QA
    UTM validationLanding-page checksPixel firingConversion-path checksApproval gates
    Paid Media Intelligence
    Budget pacingSpend anomaliesConversion volatilityCPA / ROAS movementAudience and creative performance
    SEO / GEO / AIO Intelligence
    Organic demand signalsPriority landing pagesAI Overview visibilityChatGPT / Perplexity citationsSearch and AI visibility
    Models · Claude · ACE reasoning
  5. 05

    Orchestrate

    Turn decisions into controlled workflows.

    Maintain stateRoute decisionsCoordinate agentsManage approvalsTrigger escalationHandle human handoffsDetermine which system or agent acts next
    LangGraph · n8n · ACE workflow state
  6. 06

    Act

    Execute the decision.

    ACE Agents
    Investigate anomaliesValidate launchesDiagnose tracking failuresReview search visibilityDraft client reportsRecommend optimizationsExecute approved write-backs
    Execution surfaces
    Paid mediaCRMOutboundContentReportingInternal workflows

    Client Reporting also sits here, as an output of the reasoning layer:

    Weekly summaryBusiness-impact framingIssues and actionsNext-step recommendations
  7. 07

    Learn

    Capture the result of every action and write it back into the operating graph and systems of record.

    Campaign outcomeLead progressionRevenue outcomeRecommendation accepted or rejectedIssue resolvedAgent actionHuman overridePerformance change
    Write-back → operating graph → new context → new reasoning↺ Back to 03

Blueprint, Tracking, QA, Paid Media, SEO/GEO, Reporting and Agents are ACE capabilities. Ingest through Learn is how ACE operates.