Source: 30-SUITECENTRAL-FOR-SQUIRE.md

What this source is

A business decision aid for Squire leadership (snapshot 19603d6, 2026-06-03). It does three jobs: (1) ranks the platform’s ten features by speed to a financial benefit for Squire — billable consultant-hours saved or a new billable capability on the deployments Squire already runs, not by technical depth; (2) disambiguates the two different things both called “SyncCentral”; and (3) lays out the fastest, lowest-risk path to a real return. Explicitly framed as a budget decision aid, not a code review.

Key claims

1. The one-sentence thesis: the fastest money is not a new product — it’s overlaying AI field-mapping + sync-error resolution onto Squire’s existing 40+ SyncCentral deployments, cutting consultant hours per integration. That is the Enhance path. → revenue-model, three-review-paths
2. Category differentiator = Embedded Intelligence: modules render inside the NetSuite/Business Central UI, not in a separate console (Celigo/Boomi/MuleSoft sit beside the ERP). Adoption is the rate-limiter on AI ROI, so embedded beats standalone. → embedded-intelligence, context-sidecar
3. Honest Sidecar status: the embedded surface contract, NetSuite+BC adapters, postMessage protocol, context API, and cross-system action methods all exist in code; 7 modules wired (reconciliation, lineage, approvals, sync health, compliance, flow templates, sync-error triage) — but they run against demo/in-memory data today. Pitch it as “demonstrated, pilot-ready,” not “in production.” → production-vs-demo
4. Top-10 ranking (speed-to-revenue, paraphrased): #1 Sync Error AI Assist (rides an existing error flow — fastest to bill), #2 AI Field Mapping (biggest single labor line; 95.1% top-1, 0 hallucinations, 4.9% manual-edit on a live OpenAI benchmark; Claude Haiku 4.5 scores 96.7% on the same pair, both providers 100% on a second Business Central pair), #3 Governance Pacer (enabler that makes 2 safe live), #4 Embedded Intelligence/Sidecar (premium upsell), #5 Approve-to-Apply, #6 Outbound DLP (14 patterns), #7 Per-Tenant Kill Switch, #8 Cost Transparency, #9 Record-Level Lineage, #10 Reference-Based Data Custody. → claim-proof-matrix, production-proof
5. The two “SyncCentrals” (the name-collision the corpus repeatedly trips on): Squire’s SyncCentral = the real, deployed NetSuite↔Azure integration product (~40+ deployments, live at 3PL/ShipStation + Macy’s, owner Preston). SuiteCentral 2.0’s “SyncCentral” module = a per-tenant in-memory demo surface (TenantSandbox), not real ERP traffic. SuiteCentral 2.0 wraps/upgrades the incumbent; it is not trying to be it. → squire, production-vs-demo
6. Three paths, one recommended: Integrate (overlay governance/DLP/audit — low risk, modest savings), Enhance (recommended — swap the hand-written Azure ETL transform for AI Field Mapping + Schema Drift Shield + Sync Error AI Assist; this is where the labor savings are), Replace (full SuiteApp/BC extension — not there yet). Constrained by Preston at ~20 hrs/yr on SyncCentral, so a rebuild is a non-starter. → three-review-paths, pilot-30-60-90
7. The single concrete move: pilot #1 + #2 on ONE existing SyncCentral deployment with #3 on, behind 6 controls — the smallest config that turns saved hours into a measurable number. The honest gating item before a paid pilot: a production-tier NetSuite connector test (~15 min, Preston-side). → 32-v3-netsuite-production-test-runbook

Cross-references

• Mapping accuracy 95.1% matches the live Phase-B benchmark matrix recorded in 26-canonical-metrics-and-wording and canonical-metrics — disclosed here (correctly) as a fixture figure (two ERP pairs, two providers), not a production number.
• “Enhance not Replace” is the same recommendation reached independently in 31-suitecentral-evaluation-summary — the two June-2026 leadership docs agree on path and on the single open engineering item.
• Connector honesty (5 prod / 1 beta / 11 demo / 1 stub) underpins the “demonstrated, pilot-ready” framing of the Sidecar — see production-vs-demo and preston-test-repo.

Pages updated

• revenue-model — Enhance-path “make what Squire already sells cheaper to deliver” thesis + the 40-deployment compounding argument
• embedded-intelligence — embedded-vs-standalone as the category moat and adoption-rate-limiter argument
• squire — the two-SyncCentrals disambiguation (real NetSuite↔Azure product vs. demo module)