Last updated by Errin O'Connor, Founder & Chief AI Architect, EPC Group
Standards Alignment
Control-by-control mapping of EPC Group's frameworks to NIST AI RMF, COBIT 2019, ITIL 4, DAMA-DMBOK, and CISA guidance. The skepticism killer.
See the mappingGoverned AI on Microsoft Framework
Seven layers: data classification + lineage, non-human identity, decision boundaries, escalation rules, audit trails, monitoring + kill switches, accountability mapping.
Explore the frameworkEngagement Operating Model
Seven-phase delivery methodology including Premium by Design positioning, EPC vs GSI vs Boutique comparison, and Service Standard SLAs.
Read methodologyThe EPC Group Lifecycle
Assess → Modernize → Govern → Operate → Enable. One accountable partner across the full life of a Microsoft data, analytics, and AI estate.
Explore the LifecycleGovernance-First Data Architecture
Classification, security, lineage, and compliance controls designed into the data platform before the first report ships.
Read practice8-Surface Tenant AI Readiness Standard (TAR-8)
A 24-point, evidence-based rubric for whether a Microsoft 365 and Azure tenant is ready for Copilot, agents, and external AI platforms. Eight surfaces an AI model reads — identity, sensitive data, content estate, conversations, reports and dashboards, business applications and CRM, infrastructure and external models, ownership and evidence — each scored 0 (unknown) to 3 (evidenced), with gating rules in dependency order. Version 1.0, September 2026.
Read the standardService Standard (Draft)
Published response SLAs and 90-day quick-win engagement models — high-touch, in writing. Currently draft pending sign-off.
Read draftTalk to a senior architect — not a sales rep.
Multiple models. One truth.