When an AI agent receives a complex quote request in natural language, it needs to know exactly what JSON payload structure to build. This library is not a list of UI variations; it is the ultimate menu of headless execution pathways.
Instead of building hundreds of custom quoting workflows, we map every possible way the business breathes revenue into these universal Archetypes. Once identified, we configure these specific pathways as Actions for the Agentforce LLM, bridging human intent directly to the backend ledger.
Every industry has a different conversational vocabulary. This section defines the overarching conversational boundaries the LLM must be grounded in. An agent cannot parse intent if it doesn't understand the domain's fundamental revenue mechanics.
| Domain Focus | Typical Revenue Models | LLM Cognitive Load (What the AI must understand) |
|---|---|---|
| SaaS / Cloud Services | Subscriptions, Usage, Tiers | Proration complexity, ARR/MRR consistency, and parsing mid-term upgrade/downgrade intents without losing historical data. |
| Manufacturing (B2B) | Configured BOMs, Milestones | Complex dependencies, partial shipments leading to partial billing, and tying revenue directly to delivery acceptance. |
| Telecom / Media | Bundles, Overage, Long-term | Real-time usage mediation, massive product catalogs, and high-volume amendment payloads. |
Once the LLM understands the domain, it must know how to route the data. This section dictates the execution flow—whether the AI can silently automate the payload or if it must orchestrate a human-in-the-loop approval.
| Routing Logic | What it is | Agentic Implication & API Action |
|---|---|---|
| Transactional Automation | High-volume, fast close, minimal negotiation. | Silent API Execution: The Agent parses intent, formats the payload, and dispatches directly to the order object with no human review. |
| Consultative & Guided | Diagnose, propose, and tailor to specific needs. | Draft & Approval Routing: The Agent extracts parameters to build a draft configuration, triggering a workflow for a human rep or manager to review before finalizing. |
| Usage / Consumption | Pay purely for what is consumed over time. | Telemetry Integration: The Agent's payload must establish the usage ingestion endpoints and post-period rating logic, rather than a fixed quote. |
This is where the specific data structures are defined. The LLM must know that if a user asks for a "Subscription Product," the payload strictly requires a Start Date and Term Length to be accepted by the system.
| Archetype Blueprint | Key Characteristics | Required Payload Parameters |
|---|---|---|
| One-Time (Non-Recurring) | Hardware, perpetual licenses, installation fees. | Quantity, Unit Price, Target Delivery Date. No term lengths or proration logic allowed. |
| Subscription (Termed) | SaaS licenses, support plans, maintenance. | Start Date, Term Length, Billing Frequency. Must include auto-renewal flags. |
| Hybrid Bundle (Dynamic) | Mixed fixed + variable (e.g., Base fee + overages). | Parent-Child Node Mapping. The payload must define the relational allocation between the base fee and the usage tiers. |
| Financial Adjustment | Discounts, credits, commercial corrections. | Adjustment Amount/Type, Target Asset ID. Requires strict linkage to the original order object. |
A Quote is just a promise; the fulfillment payload is the commitment. This section instructs the Agent on how to shape the downstream order data, ensuring hybrid setups are split perfectly for operations.
| Fulfillment Structure | Core Concept | RLM Downstream Impact |
|---|---|---|
| One-Time Delivery | Single instance, invoiced immediately. | Order is generated, ownership transfers immediately, and minimal lifecycle tracking is initiated. |
| Subscription-Linked Asset | Asset tied directly to a recurring timeline. | The Agent must map the Subscription to an AssetStatePeriod object, ensuring future amendments have a historical baseline. |
| Consumable / Metered | Asset consumed over time by volume. | Asset creation triggers a connection to external usage tracking systems to feed future invoices. |
This is the ultimate traceability test. It dictates exactly how the agent must format the financial data so that when it hits SAP FICO or Revenue Cloud, the ledger recognizes the money in compliance with accounting standards.
| Ledger Treatment | Core Concept | Integration Contract Rules |
|---|---|---|
| Immediate Recognition | 100% of revenue recognized at time of invoice/delivery. | Payload must hit a one-time charge type. No deferred revenue schedules can be generated. |
| Time-Based Amortization | Revenue spread evenly over the subscription term. | Payload must establish the start/end dates so the ledger can amortize daily, monthly, or quarterly automatically. |
| Multi-Element Allocation | Revenue mathematically split across a bundle. | Payload must adhere to the standalone selling price (SSP) allocation rules to satisfy audit compliance before the ledger will accept it. |