The Curriculum / Reader / Multi-Region AI Deployment
LEVEL 3 · ADVANCED · COMPANY TRACK

Multi-Region AI Deployment

This page compiles 4 files from the repository, verbatim, in reading order. The living version: this folder on GitHub.

level-3-advanced/company/05-multi-region-deployment/README.md

Multi-Region AI Deployment

Multi-region is a customer trust and continuity choice, not merely duplicated infrastructure. Design data location, model availability, latency, legal requirements, and failover behavior together.

At company scale, AI is not a side project owned by the person who writes prompts. It is a production capability spanning product, domain operations, platform engineering, security, privacy, finance, and support. Belle Realty should make the operating decision visible: what is authorized, who is accountable, what evidence is required, and how the system is stopped when reality disagrees.

Operating model

Name one directly responsible individual for the outcome and one executive sponsor for the risk. Define the decision rights for data access, model changes, vendor changes, policy exceptions, and emergency shutdowns. Maintain a registry of deployed features with purpose, customer cohorts, model and retrieval versions, data sources, tool permissions, SLOs, evaluation evidence, and review date.

Control cadence

Review leading indicators weekly: reliability, safety interventions, access denials, cost, drift, user corrections, and unresolved incidents. Review material changes before launch and at a fixed expiry date after launch. The agenda should end in decisions, owners, and dates—not a dashboard tour.

Evidence standard

Require representative offline evaluation, staged release evidence, traceable telemetry, and a documented rollback path. Segment results by geography, property type, tenant context, language, and risk level. A global pass rate is insufficient for a system that operates differently for one high-risk cohort.

Practical scenario

Before allowing an assistant to send maintenance messages across a portfolio, prove authorization boundaries, approved language, escalation behavior, provider fallback, outage handling, audit retention, and per-property cost limits. Have operations rehearse the manual path and the stop path.

Decision test

The program is mature only when a new engineer can determine what is running, why it is permitted, how it is measured, and who can change or stop it without relying on tribal knowledge.

level-3-advanced/company/05-multi-region-deployment/data-residency.md

Data Residency

Map where prompts, retrieval data, logs, backups, telemetry, support exports, and vendor subprocessors reside. Enforce regional routing and deny unsafe fallback rather than silently moving protected data.

At company scale, AI is not a side project owned by the person who writes prompts. It is a production capability spanning product, domain operations, platform engineering, security, privacy, finance, and support. Belle Realty should make the operating decision visible: what is authorized, who is accountable, what evidence is required, and how the system is stopped when reality disagrees.

Operating model

Name one directly responsible individual for the outcome and one executive sponsor for the risk. Define the decision rights for data access, model changes, vendor changes, policy exceptions, and emergency shutdowns. Maintain a registry of deployed features with purpose, customer cohorts, model and retrieval versions, data sources, tool permissions, SLOs, evaluation evidence, and review date.

Control cadence

Review leading indicators weekly: reliability, safety interventions, access denials, cost, drift, user corrections, and unresolved incidents. Review material changes before launch and at a fixed expiry date after launch. The agenda should end in decisions, owners, and dates—not a dashboard tour.

Evidence standard

Require representative offline evaluation, staged release evidence, traceable telemetry, and a documented rollback path. Segment results by geography, property type, tenant context, language, and risk level. A global pass rate is insufficient for a system that operates differently for one high-risk cohort.

Practical scenario

Before allowing an assistant to send maintenance messages across a portfolio, prove authorization boundaries, approved language, escalation behavior, provider fallback, outage handling, audit retention, and per-property cost limits. Have operations rehearse the manual path and the stop path.

Decision test

The program is mature only when a new engineer can determine what is running, why it is permitted, how it is measured, and who can change or stop it without relying on tribal knowledge.

level-3-advanced/company/05-multi-region-deployment/failover-strategy.md

Regional Failover Strategy

Predefine whether failover stays in region, degrades functionality, queues work, or requires consent. Rehearse provider and region failures; an untested cross-region path is an assumption, not resilience.

At company scale, AI is not a side project owned by the person who writes prompts. It is a production capability spanning product, domain operations, platform engineering, security, privacy, finance, and support. Belle Realty should make the operating decision visible: what is authorized, who is accountable, what evidence is required, and how the system is stopped when reality disagrees.

Operating model

Name one directly responsible individual for the outcome and one executive sponsor for the risk. Define the decision rights for data access, model changes, vendor changes, policy exceptions, and emergency shutdowns. Maintain a registry of deployed features with purpose, customer cohorts, model and retrieval versions, data sources, tool permissions, SLOs, evaluation evidence, and review date.

Control cadence

Review leading indicators weekly: reliability, safety interventions, access denials, cost, drift, user corrections, and unresolved incidents. Review material changes before launch and at a fixed expiry date after launch. The agenda should end in decisions, owners, and dates—not a dashboard tour.

Evidence standard

Require representative offline evaluation, staged release evidence, traceable telemetry, and a documented rollback path. Segment results by geography, property type, tenant context, language, and risk level. A global pass rate is insufficient for a system that operates differently for one high-risk cohort.

Practical scenario

Before allowing an assistant to send maintenance messages across a portfolio, prove authorization boundaries, approved language, escalation behavior, provider fallback, outage handling, audit retention, and per-property cost limits. Have operations rehearse the manual path and the stop path.

Decision test

The program is mature only when a new engineer can determine what is running, why it is permitted, how it is measured, and who can change or stop it without relying on tribal knowledge.

level-3-advanced/company/05-multi-region-deployment/regional-model-selection.md

Regional Model Selection

Select approved models per region using capability, latency, data terms, language coverage, safety evidence, and capacity. Model choice is a policy decision that should be encoded in the router.

At company scale, AI is not a side project owned by the person who writes prompts. It is a production capability spanning product, domain operations, platform engineering, security, privacy, finance, and support. Belle Realty should make the operating decision visible: what is authorized, who is accountable, what evidence is required, and how the system is stopped when reality disagrees.

Operating model

Name one directly responsible individual for the outcome and one executive sponsor for the risk. Define the decision rights for data access, model changes, vendor changes, policy exceptions, and emergency shutdowns. Maintain a registry of deployed features with purpose, customer cohorts, model and retrieval versions, data sources, tool permissions, SLOs, evaluation evidence, and review date.

Control cadence

Review leading indicators weekly: reliability, safety interventions, access denials, cost, drift, user corrections, and unresolved incidents. Review material changes before launch and at a fixed expiry date after launch. The agenda should end in decisions, owners, and dates—not a dashboard tour.

Evidence standard

Require representative offline evaluation, staged release evidence, traceable telemetry, and a documented rollback path. Segment results by geography, property type, tenant context, language, and risk level. A global pass rate is insufficient for a system that operates differently for one high-risk cohort.

Practical scenario

Before allowing an assistant to send maintenance messages across a portfolio, prove authorization boundaries, approved language, escalation behavior, provider fallback, outage handling, audit retention, and per-property cost limits. Have operations rehearse the manual path and the stop path.

Decision test

The program is mature only when a new engineer can determine what is running, why it is permitted, how it is measured, and who can change or stop it without relying on tribal knowledge.

← Red-Team Program Privacy and Compliance for AI →