Data-layer swarm · Inventory-orchestration agent · Cross-channel-action-coordination skill · Build pillar · Published July 11, 2026

How to build cross-channel action coordination for multi-location ecom + retail

Multi-location multi-channel operators running ecom + retail + marketplace surfaces work above a strong IMS + OMS + commerce + POS + marketplace + ad-platform + lifecycle primitives layer (NetSuite + Cin7 + Brightpearl + Linnworks + ChannelAdvisor + Sellbrite + Skubana + Veeqo + Ordoro + Ecomdash + Sellbery for IMS/OMS; Shopify + BigCommerce + WooCommerce + Magento + Adobe Commerce + Salesforce Commerce Cloud for commerce; Square + Toast + Lightspeed + Clover + Aloha NCR for POS; Amazon SP-API + Walmart Marketplace + Target Plus + eBay + Etsy + Faire for marketplace; Google Ads + Meta Marketing + TikTok Marketing + Microsoft Ads for ad platforms; Klaviyo + Mailchimp + Iterable + Braze for lifecycle). The orchestration that sits above those primitives — a per-SKU per-location inventory stream that normalizes across the operator source surface, a per-channel action map that routes inventory-state changes to operator- controlled actions, a per-inventory-state auto-pause engine with policy-as-code gating, a per-backorder auto-resume engine with warmup, a per-cross-channel reconciliation layer, a multi- LLM classification layer, a feedback loop, and a per-action compliance gate that ties decisions to FTC MARS, FTC substantiation, Robinson-Patman, FDD Item 12/17/19, and Lanham Act anchors — is operator-side architecture. This guide explains how to architect the cross-channel-action-coordination skill on the inventory-orchestration agent end-to-end.

Start with the AI Readiness Assessment Explore Fractional CMO with AI Swarm Take the 3-question fit quiz

What you will build

A per-SKU per-location inventory stream via CDC capture (Debezium, AWS DMS, Striim, Fivetran, HVR, Qlik Replicate) plus webhook ingestion plus polling fallback, with operator-configured cadence (real-time, 1-minute, 5-minute, hourly), per-source rate-limit and retry handling, dead-letter queues, schema-evolution handling, deduplication, and canonicalization. Eight event classes: inventory delta, stockout, back-in-stock, low-stock threshold, allocation change, receipt, shipment, return.
A per-channel action map that routes inventory-state changes to operator-controlled actions across ad platforms (Google Ads, Meta Marketing API, TikTok Marketing, Microsoft Ads pause/resume), lifecycle vendors (Klaviyo, Mailchimp, Iterable, Braze segment suppression), PDP state (out-of-stock state), landing-page state, GBP product visibility, and Shopify product status, with cross-channel action precedence rules.
A per-inventory-state auto-pause engine with six triggers (out-of-stock now, out-of-stock imminent in 1 hour, out-of-stock imminent in 24 hours, low stock 7-day, allocation locked, backorder imminent) encoded as policy-as- code (OPA Rego, AWS Cedar, Casbin, Cerbos, Oso), with blast- radius estimation, cost-impact estimation, and explainability surface.
A per-backorder auto-resume engine with four confidence classes (confirmed receipt, allocation released, reorder pending with confirmed vendor ship date, projected back-in-stock from the demand-forecast handoff), grace-period rules to prevent premature resumes on flapping inventory, and warmup logic (budget ramp, bid ramp, audience warmup).
A per-cross-channel reconciliation layer(cross-vendor merge, deduplication, conflict resolution with stricter-of-both, precedence rules, final-state attestation).
A multi-LLM classification layer (operator chooses across OpenAI, Anthropic, Google, Mistral, Cohere, Meta) with trigger classification, blast-radius extraction, cost/revenue-impact extraction, confidence scoring, self- consistency cross-check, and chain-of-thought extraction.
Per-action confidence intervals and explainability (Bayesian posterior, frequentist confidence, bootstrap confidence, confidence tier; SHAP, LIME, feature importance, counterfactual explanations, attribution graph, causal DAG, multi-LLM narrative).
A feedback loop comparing realized vs predicted action outcomes (did the pause actually save the projected ad-spend, did the resume actually capture the projected revenue) and recalibrating blast-radius, cost- impact, and confidence models with false-positive and false- negative pattern learning.
A per-action compliance gate anchored on FTC MARS multi-location/multi-channel claim consistency, FTC substantiation doctrine, Robinson-Patman Act, FDD Item 12/17/19, and Lanham Act, extended to state subscription auto-renewal + ROSCA + Click-to-Cancel + CARD Act + CCPA/CPRA + GDPR + EU AI Act + NIST AI RMF + ISO 42001/27001 + SOC 2 Type II + PCI DSS 4.0 + Sarbanes-Oxley + FCC TCPA via policy-as-code (OPA Rego, AWS Cedar, Casbin, Cerbos, Oso) that operator counsel reviews.
Cross-skill handoffs and an audit trail to siblings on the inventory-orchestration agent and broader swarm, with audit trail to operator-controlled WORM storage at per-statute retention windows operator counsel sets.

Where the orchestration above IMS, commerce, marketplace, and ad- platform primitives compounds at multi-location multi-channel scale

The vendor primitives are strong. IMS and OMS vendors ship per-account inventory sync. Commerce platforms ship per-SKU availability broadcasting. POS platforms ship per-store sale and receipt events. Marketplace channels ship per-listing availability. Ad platforms ship per-campaign pause/resume. Lifecycle vendors ship per-segment suppression. The orchestration above those primitives is what compounds at multi-location multi-channel scale.

The first operationally distinctive constraint is FTC MARS multi-location/multi-channel claim consistency. When the same SKU shows in-stock on one channel and out-of-stock on another at the same geographic moment, or when back-in-stock claims drift across channels, FTC scrutiny of claim consistency applies. The per-action gate routes per-channel inventory drift to operator-counsel-reviewed reconciliation.

The second distinctive constraint is FTC substantiation doctrine (Pfizer 1972 plus the broader Reasonable-Basis Doctrine) applied to availability claims. In-stock, ships- from-store, ready-for-pickup, back-in-stock-soon, and restocking-in-N-days are advertising claims; the gate ties every channel update to the underlying inventory-state evidence and records the substantiation linkage.

The third distinctive constraint is Robinson-Patman Act and FTC 2024-2025 reinvigorated price-discrimination enforcement. Cross-banner and cross-channel availability and pricing decisions that vary based on banner, channel, or trade area expose Robinson-Patman analysis when functional-equivalent- availability defenses fall short. The gate captures the substantiation record at the decision moment.

The fourth distinctive constraint is FDD Item 12 (territorial protection), Item 17 (renewal, termination, and transfer), and Item 19 (financial-performance representation) when cross-channel actions affect franchisee revenue. Channel- action coordination that shifts revenue between franchisees routes to operator-counsel-reviewed FDD-aware workflows.

The fifth distinctive constraint is Lanham Act false advertising for comparative and competitive-availability claims. When cross-channel updates create comparative or competitive-availability inconsistencies, Lanham Act exposure can attach. The gate emits the comparative-claim evidence and routes to counsel review.

Beyond the five anchors, the gate also covers state subscription auto-renewal laws (when back-in-stock auto- resume triggers subscription billing); ROSCA + FTC Click-to- Cancel + state CARD Act analogs; California Automatic Renewal Law; CCPA/CPRA + GDPR Article 22 + PIPEDA + CASL + LGPD + DPDP when action coordination joins customer identity for personalization; EU AI Act Articles 13/14 when the auto- pause/auto-resume engine drives automated decisioning; EU DSA Article 26 + EU DMA platform-side obligations; NIST AI RMF + ISO 42001 + ISO 27001 + SOC 2 Type II; PCI DSS 4.0 when payment-step inventory holds; Sarbanes-Oxley Section 302/404 for inventory and revenue accounting controls; FCC TCPA + 10DLC when lifecycle suppression touches SMS. The gate is policy-as-code (OPA Rego, AWS Cedar, Casbin, Cerbos, Oso); operator counsel reviews rule updates.

The real ecosystem the orchestration sits above

IMS, OMS, and POS primitives

NetSuite, Cin7, Brightpearl, Linnworks, ChannelAdvisor (Rithum), Sellbrite, Skubana (Extensiv), Veeqo, Ordoro, Ecomdash, Sellbery for IMS/OMS; Square, Toast, Lightspeed, Clover, Aloha NCR for POS. Strong primitives for per-account inventory sync and per-store sale/receipt events. The per-SKU per-location inventory stream sits above this layer.

Commerce and marketplace primitives

Shopify, BigCommerce, WooCommerce, Magento, Adobe Commerce, Salesforce Commerce Cloud for commerce; Amazon SP-API, Walmart Marketplace, Target Plus, eBay, Etsy, Faire for marketplace. Strong primitives for per-SKU availability broadcasting and per-listing availability. The per-channel action map sits above this layer with cross-channel precedence rules.

Ad-platform and lifecycle primitives

Google Ads, Meta Marketing API, TikTok Marketing, Microsoft Ads for ad platforms; Klaviyo, Mailchimp, Iterable, Braze for lifecycle. Strong primitives for per-campaign pause/resume and per-segment suppression. The auto-pause/auto-resume engines route to these endpoints under policy-as-code gating.

CDC, integration, and compliance-tooling primitives

Debezium, AWS DMS, Striim, Fivetran, HVR, Qlik Replicate for CDC; Temporal, AWS Step Functions, Apache Airflow, Dagster, Prefect for workflow orchestration; Hyperproof, Drata, Vanta, Thoropass for SOC 2 / ISO control evidence; OneTrust, TrustArc, Ketch, Securiti, BigID for privacy program tooling. Strong primitives. The per-action compliance overlay coordinates them via a policy-as-code gate (OPA Rego, AWS Cedar, Casbin, Cerbos, Oso) that operator counsel reviews.

How the architecture is built

Inventory-stream substrate. Subscribe to IMS, OMS, commerce, POS, and marketplace webhooks. Wire CDC via Debezium, AWS DMS, Striim, Fivetran, HVR, or Qlik Replicate. Configure polling fallback. Land events in the operator data warehouse (Snowflake, Databricks, BigQuery, Redshift, Postgres) at the per-SKU per-location canonical-ID grain. Classify events into the 8-class taxonomy.
Per-channel action map. Wire pause/resume, suppression, PDP-state, landing-page-state, GBP product visibility, and Shopify product status endpoints. Encode cross-channel precedence rules. Validate per-channel authorization scope.
Per-inventory-state auto-pause engine. Encode the six triggers as policy-as-code on OPA Rego, AWS Cedar, Casbin, Cerbos, or Oso. Compute blast-radius and cost-impact estimates. Emit explainability records.
Per-backorder auto-resume engine. Consume back-in-stock events with confidence classes. Apply operator- counsel-reviewed resume policy-as-code. Apply grace periods against flapping inventory. Apply warmup logic for budget, bid, and audience ramp.
Cross-channel reconciliation. Run cross-vendor merge, deduplication, conflict resolution with stricter-of- both, precedence rules, and final-state attestation.
Multi-LLM classification. Ensemble multiple vendor LLM APIs for trigger classification, blast-radius extraction, cost-impact extraction, revenue-impact extraction, confidence scoring, self-consistency cross-check, and chain- of-thought extraction.
Per-action confidence + explainability.Compute Bayesian posterior, frequentist, and bootstrap confidence intervals. Surface SHAP, LIME, feature importance, counterfactual, attribution-graph, causal-DAG, and multi-LLM narrative explainability.
Feedback loop. Compare realized vs predicted action outcomes. Recalibrate blast-radius, cost-impact, and confidence models. Learn false-positive and false-negative patterns.
Per-action compliance gate. Express the gate as policy-as-code on OPA Rego, AWS Cedar, Casbin, Cerbos, or Oso. Encode the five distinctive anchors (FTC MARS, FTC substantiation, Robinson-Patman, FDD Item 12/17/19, Lanham Act) plus the broader compliance surface. Operator counsel reviews every rule update.
Cross-skill handoffs. Hand off to siblings on the inventory-orchestration agent (multi-channel inventory management, inventory-aware PDP variation, inventory visibility, SEO-preserving PDP refresh, product feed management, multi-PIM catalog ingestion, marketplace management software, PIM software, creative real-estate API, POS integration) and across the broader swarm (marketing mix modeling, PPC budget management, per-market budget recommendations, auto-remediation for low-risk drift, real- time multi-location inventory state monitoring, BOPIS friction detection, versioned product history for recall traceability).
Audit trail. Emit a per-action canonical audit record to operator-controlled WORM storage (AWS S3 Object Lock, GCS retention, Azure Blob immutable, Snowflake Time Travel) with per-statute retention windows operator counsel sets (IRS 7yr, FTC 7yr, SEC 6yr, SOX 7yr, CFPB Reg E 3yr).

Frequently asked

What does cross-channel action coordination do that a single IMS account-level inventory sync does not?

Inventory and order management vendors (NetSuite, Cin7, Brightpearl, Linnworks, ChannelAdvisor Rithum, Sellbrite, Skubana Extensiv, Veeqo, Ordoro, Ecomdash, Sellbery) ship strong primitives for per-account per-warehouse per-SKU inventory sync. Commerce platforms (Shopify, BigCommerce, WooCommerce, Magento, Adobe Commerce, Salesforce Commerce Cloud) ship strong primitives for per-SKU availability broadcasting. POS platforms (Square, Toast, Lightspeed, Clover, Aloha NCR) ship strong primitives for per-store sale and receipt events. Marketplace channels (Amazon SP-API, Walmart Marketplace, Target Plus, eBay, Etsy, Faire) ship strong primitives for per-listing availability. Ad platforms (Google Ads, Meta Marketing API, TikTok Marketing, Microsoft Ads) ship strong primitives for per-campaign pause/resume. Lifecycle vendors (Klaviyo, Mailchimp, Iterable, Braze) ship strong primitives for per-segment suppression. Cross-channel action coordination sits above this layer for multi-location multi-channel operators, and adds: a per-SKU per-location inventory stream that joins CDC capture, webhook ingestion, and polling fallback across all source systems with operator-set cadence (real-time, 1-minute, 5-minute, hourly), rate-limit and retry handling, dead-letter queues, schema-evolution handling, deduplication, canonicalization, and event classification (inventory delta, stockout, back-in-stock, low-stock threshold, allocation change, receipt, shipment, return); a per-channel action map that routes inventory-state changes to operator-controlled actions (Google Ads pause/resume, Meta Marketing pause/resume, TikTok Marketing pause/resume, Microsoft Ads pause/resume, Klaviyo segment suppression, Mailchimp suppression, Iterable suppression, Braze suppression, PDP out-of-stock state, landing-page state, GBP product visibility, Shopify product status) with cross-channel action precedence rules; a per-inventory-state auto-pause engine with triggers (out-of-stock now, out-of-stock imminent in 1 hour, out-of-stock imminent in 24 hours, low stock 7-day, allocation locked, backorder imminent) encoded as policy-as-code (OPA Rego, AWS Cedar, Casbin, Cerbos, Oso), blast-radius estimation, cost-impact estimation, and explainability surface; a per-backorder auto-resume engine with back-in-stock confidence classes (confirmed receipt, allocation released, reorder pending, projected back-in-stock), grace-period rules, and warmup logic (budget ramp, bid ramp, audience warmup); a per-cross-channel reconciliation layer (cross-vendor merge, deduplication, conflict resolution with stricter-of-both, precedence rules, final-state attestation); a multi-LLM classification layer (operator chooses across OpenAI, Anthropic, Google, Mistral, Cohere, Meta) with trigger classification, blast-radius extraction, cost/revenue-impact extraction, confidence scoring, self-consistency cross-check, and chain-of-thought extraction; per-action confidence intervals (Bayesian posterior, frequentist confidence, bootstrap confidence, confidence tier); per-action explainability (SHAP, LIME, feature importance, counterfactual explanations, attribution graph, causal DAG, multi-LLM narrative); a feedback loop comparing realized vs projected action outcomes and recalibrating blast-radius, cost-impact, and confidence models; a per-action compliance gate (covered in the next answer); and an audit trail to operator-controlled WORM storage at per-statute retention windows.

What are the operationally distinctive compliance anchors for cross-channel action coordination, and how does the per-action compliance gate cover them?

Five anchors sit at the operational center of cross-channel action coordination that off-the-shelf commerce compliance overlays often miss. Anchor 1 — FTC MARS multi-location/multi-channel claim consistency. When the same SKU shows in-stock on one channel and out-of-stock on another at the same geographic moment, or when back-in-stock claims drift across channels, FTC scrutiny of claim consistency applies. The per-action gate routes per-channel inventory drift to operator-counsel-reviewed reconciliation and emits a substantiation record at the moment of every availability claim. Anchor 2 — FTC substantiation doctrine (Pfizer 1972 plus the broader Reasonable-Basis Doctrine) for availability claims. In-stock, ships-from-store, ready-for-pickup, back-in-stock-soon, and restocking-in-N-days are advertising claims that must be substantiable at the moment of publication. The per-action gate ties every channel update to the underlying inventory-state evidence and records the substantiation linkage. Anchor 3 — Robinson-Patman Act and FTC 2024-2025 reinvigorated price-discrimination enforcement. Cross-banner and cross-channel availability and pricing decisions that vary based on banner, channel, or trade area expose Robinson-Patman analysis when functional-equivalent-availability defenses fall short. The per-action gate captures the substantiation record at the decision moment and routes per-banner/per-channel variation to counsel-reviewed workflows. Anchor 4 — FDD Item 12 (territorial protection), Item 17 (renewal/termination/transfer), and Item 19 (financial-performance representation) when cross-channel actions affect franchisee revenue. Channel-action coordination that shifts revenue between franchisees (pause on franchisee-A geo while resuming on franchisee-B geo) routes to operator-counsel-reviewed FDD-aware workflows. Anchor 5 — Lanham Act false advertising for comparative and competitive-availability claims. When cross-channel updates create comparative or competitive-availability inconsistencies, Lanham Act exposure can attach. The per-action gate emits the comparative-claim evidence and routes to counsel review. Beyond the five anchors, the per-action gate also covers state subscription auto-renewal laws (when back-in-stock auto-resume triggers subscription billing); ROSCA + FTC Click-to-Cancel + state CARD Act analogs; California Automatic Renewal Law; CCPA/CPRA + GDPR Article 22 + PIPEDA + CASL + LGPD + DPDP when action coordination joins customer identity for personalization; EU AI Act Articles 13/14 when the auto-pause/auto-resume engine drives automated decisioning customers see; EU DSA Article 26 + EU DMA platform-side obligations; NIST AI RMF + ISO 42001 + ISO 27001 + SOC 2 Type II; PCI DSS 4.0 when payment-step inventory holds; Sarbanes-Oxley Section 302/404 for inventory and revenue accounting controls; FCC TCPA + 10DLC when lifecycle suppression touches SMS. The gate is policy-as-code on OPA Rego, AWS Cedar, Casbin, Cerbos, or Oso, with operator counsel reviewing rule updates.

How do the per-SKU per-location inventory stream, per-channel action map, and per-inventory-state auto-pause engine actually work?

The per-SKU per-location inventory stream joins CDC capture (via Debezium, AWS DMS, Striim, Fivetran, HVR, Qlik Replicate) plus webhook subscriptions plus polling fallback across the source surface. Per-source cadence (real-time, 1-minute, 5-minute, hourly) is operator-configured against per-vendor API rate limits. Per-source retry and dead-letter queue handling cover transient failures. Schema-evolution handling tolerates vendor-side schema changes. Per-SKU per-location stream deduplication and canonicalization yield a single canonical event stream with classified events (inventory delta, stockout, back-in-stock, low-stock threshold, allocation change, receipt, shipment, return). The per-channel action map routes inventory-state changes to operator-controlled actions across the channel surface (Google Ads pause/resume, Meta Marketing pause/resume, TikTok Marketing pause/resume, Microsoft Ads pause/resume, Klaviyo segment suppression, Mailchimp suppression, Iterable suppression, Braze suppression, PDP out-of-stock state, landing-page state, GBP product visibility, Shopify product status). Cross-channel action precedence rules resolve conflicts when multiple actions target the same SKU on the same channel. The per-inventory-state auto-pause engine fires on operator-defined triggers (out-of-stock now, out-of-stock imminent in 1 hour, out-of-stock imminent in 24 hours, low stock 7-day, allocation locked, backorder imminent), evaluates the policy-as-code gate (OPA Rego, AWS Cedar, Casbin, Cerbos, Oso) for permission, estimates the blast radius (which campaigns, segments, PDPs, GBP listings, and Shopify product entries are affected), estimates the cost impact (ad-spend at risk, revenue at risk, channel-disapproval risk), surfaces an explainability record, and feeds the action through the feedback loop.

How does the per-backorder auto-resume engine, cross-channel reconciliation, and cross-skill handoff coordinate with the rest of the swarm?

The per-backorder auto-resume engine consumes back-in-stock events with confidence classes (confirmed receipt at the receiving dock, allocation released by the allocation engine, reorder pending with confirmed vendor ship date, projected back-in-stock from the demand-forecast handoff) and applies operator-counsel-reviewed resume policy-as-code. A grace period prevents premature resumes on flapping inventory. A warmup phase ramps budget (budget ramp), bid (bid ramp), and audience (audience warmup) so that the resumed campaign does not pour spend into a still-fragile inventory pool. The per-cross-channel reconciliation layer runs cross-vendor merge across operator-controlled per-channel state sources, deduplication of duplicate signals, conflict resolution via stricter-of-both, precedence rules operator counsel and ops jointly set, and final-state attestation emitted to the audit trail. The multi-LLM classification layer ensembles multiple vendor LLM APIs for trigger classification, blast-radius extraction, cost-impact extraction, revenue-impact extraction, confidence scoring, self-consistency cross-check, and chain-of-thought extraction. The feedback loop compares realized vs predicted action outcomes (did the pause actually save the projected ad-spend, did the resume actually capture the projected revenue) and recalibrates blast-radius, cost-impact, and confidence models with false-positive and false-negative pattern learning. The skill hands off to siblings on the inventory-orchestration agent (multi-channel inventory management, inventory-aware PDP variation, inventory visibility, SEO-preserving PDP refresh, product feed management, multi-PIM catalog ingestion, marketplace management software, PIM software, creative real-estate API, POS integration) and across the broader swarm (marketing mix modeling, PPC budget management, per-market budget recommendations, auto-remediation for low-risk drift, real-time multi-location inventory state monitoring, BOPIS friction detection, versioned product history for recall traceability).

What does Completions report on a Tier 3 engagement that covers cross-channel action coordination?

Tier 3 engagements report against a pre-engagement baseline that the Tier 1 assessment establishes for the operator stack. The reporting cycle covers six workstreams: (1) per-SKU per-location inventory stream coverage observed across the operator IMS + OMS + commerce + POS + marketplace surface, with per-source ingestion completeness and CDC lag reported; (2) per-channel action map coverage observed across the operator ad-platform + lifecycle + PDP + GBP surface, with per-channel action latency and disapproval-detection diagnostics reported; (3) per-inventory-state auto-pause surface observed against operator-counsel-reviewed trigger and policy definitions, with blast-radius and cost-impact diagnostics reported per cohort; (4) per-backorder auto-resume surface observed across the confidence-class layer and warmup phase, with grace-period and false-resume diagnostics reported; (5) cross-channel reconciliation surface observed across cross-vendor merge, deduplication, and precedence-rule flows, with conflict-resolution diagnostics reported; (6) per-action compliance gate pass rate observed across FTC MARS + FTC substantiation + Robinson-Patman + FDD Item 12/17/19 + Lanham Act + state subscription auto-renewal + CCPA/CPRA + GDPR + EU AI Act + NIST AI RMF + ISO 42001/27001 + SOC 2 Type II + PCI DSS 4.0 + Sarbanes-Oxley + FCC TCPA scope. Caveats: IMS/OMS/commerce/POS/marketplace vendor API rate limits, CDC vendor lag, ad-platform pause/resume cadence, lifecycle-vendor suppression cadence, LLM-vendor availability, and per-statute retention windows shifting with operator counsel policy sit outside Completions control and are reported alongside observed performance; attorney-client privilege on counsel-reviewed substantiation records and reconciliation rules is preserved through every layer. Completions does not commit to fixed numeric SLAs on stream coverage, action map coverage, auto-pause precision, auto-resume confidence, reconciliation accuracy, or compliance pass rate when those KPIs depend on vendor performance, marketplace decisioning, or counsel policy decisions.

Engage Completions

Start with the AI Readiness Assessment (Tier 1, 2-3 weeks). If the operation is ready to absorb the cross-channel- action-coordination skill on the inventory-orchestration agent, the assessment hands off to the AI Swarm Setup Sprint (Tier 2, 4-8 weeks). If the operation needs ongoing orchestration after Tier 2 hand-off, the skill continues under Fractional CMO with AI Swarm (Tier 3, 6-month minimum, 1-2 days/wk embedded). Operator owns every artifact at every tier. Operator can in-house at any time.