IIVPA/Resources/Knowledge base

Knowledge base

Reference articles, use cases and guides on AI agents in the enterprise — organised by industry, team and system.

48articles 26categories Oct 2026last updated
K
Industries
Teams
Systems
48 articles
Sort
KB-1000 Guest requests across channels Understanding WhatsApp, email and chat requests and actioning them in the PMS. Use case HospitalityCustomer serviceHotel PMS 08 Oct 20263 min KB-1037 Group and event enquiries Reading RFPs, checking availability and drafting proposals. Use case HospitalitySalesMicrosoft 365 03 Sep 202610 min KB-1074 Night audit and reconciliation Reconciling folios, payments and channel reports overnight. Use case HospitalityFinanceHotel PMS 27 Sep 20268 min KB-1111 Supplier invoice matching Matching F&B and housekeeping invoices to POs and receipts. Use case HospitalityFinanceSAP 14 Aug 20266 min KB-1148 Order and return queries Handling status, change and return requests end to end. Use case RetailCustomer serviceSalesforce 11 Sep 20264 min KB-1185 Product content preparation Turning supplier data into consistent listings. Use case RetailOperationsMicrosoft 365 02 Oct 202611 min KB-1222 Store incident triage Classifying and routing store incident reports. Use case RetailOperationsServiceNow 26 Aug 20269 min KB-1259 Supplier compliance checks Checking certificates and specifications against policy. Use case RetailProcurementSAP 19 Sep 20267 min KB-1296 KYC and onboarding files Assembling and screening due-diligence files. Use case BankingLegal & complianceMicrosoft 365 08 Oct 20265 min KB-1333 Trade finance document checks Checking letters of credit against rules. Use case BankingOperationsOracle 03 Sep 20263 min KB-1370 Regulatory change mapping Mapping new circulars to internal policies. Use case BankingLegal & complianceMicrosoft 365 27 Sep 202610 min KB-1407 Payment investigations Researching payment exceptions with evidence. Use case BankingOperationsOracle 14 Aug 20268 min

Glossary

10 terms
AI agent Software that reads, decides and acts across systems to complete a defined task, and asks a person when uncertain.
Audit trail A complete, tamper-evident record of what an agent received, decided and did.
Autonomy level The degree to which an agent may act without human approval.
Evaluation Testing an agent on real historical cases before and after each change.
Guardrail A rule limiting what an agent may do, enforced outside the AI model.
Hallucination An AI output not supported by the source information. Mitigated by citations and checks.
Human in the loop A design in which a person reviews or approves agent work at defined points.
Integration A secure connection allowing an agent to read from or write to a system.
Orchestration Coordinating several specialist agents within one process.
RPA Robotic process automation: bots that follow fixed, screen-level rules.