Loading…
Loading…
Refund decisions involve customer experience risk and financial exposure. Archiet generates a governed refund and retention agent: the DMN policy table encodes your support policy, the BPMN workflow manages the approval and retention offer flow, and the AI only reads case history — it never decides the outcome.
Refund decisions combine financial exposure with customer experience risk. Approve too readily and you attract refund abuse; deny too strictly and you accelerate churn — especially among your most valuable customers. Support agents make inconsistent decisions because the policy is not explicit. Managers override decisions arbitrarily. Neither path is auditable. A governed refund agent solves both problems: the DMN policy table makes the policy explicit and consistent, the audit trail makes every decision reviewable, and the retention offer logic runs deterministically — no support agent discretion required for standard cases.
The agent implements four decision factors. Billing error: if the refund reason is a billing error confirmed by the system, the refund is auto-approved regardless of amount or tenure — no discretion should be exercised on a company mistake. Account tenure: customers under 90 days are outside the standard refund window except for billing errors. Long-tenure customers (24+ months) receive automatic approval with a retention offer. VIP customers receive expedited approval. Amount thresholds: standard refunds below policy limits auto-approve; large refunds escalate to a support manager. The BPMN workflow handles the retention offer presentation — if a customer is approved for a refund and also qualifies for a retention incentive (credit, discount, plan upgrade), the BPMN sends both simultaneously.
The refund agent is not just a binary approve/deny machine — it integrates retention offers as a first-class output. When the DMN policy indicates a long-tenure customer requesting a refund within the standard amount range, the output includes both the refund approval and a retention offer payload (e.g., 30% discount on next month, plan upgrade trial, or account credit). The BPMN process sends the retention offer alongside the refund confirmation. Support agents see both outcomes in the operator console and can track which customers accepted the retention offer vs. took the refund and churned.
Every refund decision creates an audit record: customer segment, tenure, refund reason, amount, DMN rule matched, outcome (approved/denied/partial), and whether a retention offer was made and accepted. The operator console aggregates this into a refund analytics dashboard: approval rate by reason category, override rate by support manager (a high override rate signals the policy needs adjustment), retention offer acceptance rate, and churn rate for customers who received a refund. These metrics are derived from the audit trail — no separate analytics pipeline required.
Yes. The DMN table includes reason_category as an input field. Standard categories: billing_error (highest approval priority), dissatisfied (tenure-based), competitor (flag for retention offer), other (standard process). Support analysts can add custom reason categories and configure specific policy rules for each without a code change.
The DMN table supports a product_type field (subscription, one-time, seat-based) that can drive different policy rules. Subscription refunds can be prorated automatically; one-time purchase refunds apply the standard window; seat-based products can have per-seat refund logic. The product_type is extracted from the customer's account record, not from the refund request itself.
Yes, with a logged audit trail. The operator console includes an override button that requires the manager to select a reason. Overrides are recorded in the audit trail with the manager's identity, the original decision, the override decision, and the stated reason. Override rates by manager and by reason category are visible in the analytics dashboard — a leading indicator of policy gaps or training needs.
CRM refund automation is typically a simple rule engine without an audit trail or retention integration. The governed agent adds: a formal DMN policy table (exportable, versionable, portable), a BPMN workflow that integrates retention offers, a complete audit trail for every decision, and the ability to deploy as a standalone service that any customer-facing system can call. You own the policy logic; it is not locked inside the CRM vendor's proprietary rule engine.
Most AI invoice automation is a black box — the model decides, you can't explain why. Archiet generates a governed invoice approval agent: the approval flow is BPMN (auditable), the decision logic is a DMN policy table your finance team can edit without code changes, and the AI only reads documents — it never decides the outcome.
Enterprise procurement teams need approval automation that can be explained to auditors, edited by business analysts, and trusted in regulated industries. Archiet generates a governed procurement agent: BPMN workflow, DMN policy table, bounded LLM for document reading, and a complete audit trail from day one.
Insurance claims triage requires speed, accuracy, and full auditability. Archiet generates a governed claims triage agent: FNOL intake, AI-powered data extraction from claim documents, deterministic DMN severity routing that actuaries control, and a complete audit trail for regulatory examination.
free plan. No credit card required. Generate your first compliant architecture blueprint in under 10 minutes.