Concepts & glossary
The core vocabulary of the LendEasy platform — case, task, queue, interaction, action request, compliance decision, promise-to-pay, evidence, governed command, payoff projection, form entry, and reconciliation exception.
These are the terms that recur across the guides and the API. They mean the same thing everywhere — in the product, in these docs, and in the audit trail.
Core concepts
| Term | Meaning |
|---|---|
| Case | The durable record of one customer or merchant matter — a question, delinquency, bankruptcy notice, cease request, dispute, or operational issue. It owns why the matter exists and what must be true before it closes. |
| Task | A single unit of work, created by a workflow, an interaction, a case control, a system monitor, an AI finding, or an action result. It reaches an eligible worker through a queue. |
| Queue | A prioritized container that routes work to eligible workers by task type, skill, channel, case type, priority, or restriction. |
| Interaction | An inbound, outbound, or externally logged communication with a customer or authorized party. It is distinct from a task, provider event, and resulting action. |
| Action request | A case-linked request for your system of record to perform or record an action it owns, such as a fee waiver, a reschedule, or a payment. LendEasy owns the request lifecycle; the core owns the authoritative result. |
| Compliance decision | The structured result of evaluating a proposed contact or action against policy and known facts: allowed, allowed with warning, approval required, blocked, or missing facts. |
| Promise to pay | A customer commitment with one or more amounts and dates, evaluated only against authoritative posted payment facts. It never moves money by itself. |
| Evidence | Material that supports a decision or outcome: documents, interactions, notes, snapshots from the system of record, compliance decisions, and AI analysis. |
| Worker | A human agent, an AI agent, or a system monitor. All three perform or prepare work under the same permissions, compliance, and audit. |
| Work packet | The bounded context, policy, allowed tools, prohibited actions, and approval path assembled for a worker on a task. An AI agent sees and does only what its packet grants. |
| Autonomy | The action-specific level an AI may observe, assist, prepare, or execute within policy. It is capped by risk and earned by version-specific evidence. |
Lending Core terms
The Lending Core under /v1 has vocabulary of its own. Use these terms exactly; older phrasings — “loan actions,” “payoff quotes,” “form submissions,” “reconciliation items” — do not name anything in the API.
| Term | Meaning |
|---|---|
| Governed command | A ledger operation — interest waiver, write-off, charge-off, re-age, re-amortize, and their undos — executed as POST /v1/loans/{loanId}/transactions?command=… under case authority, governance records, and maker-checker. |
| Command envelope | The { resourceId, changes } body a Lending Core mutation returns. Read the resource afterward for its canonical state. |
| Payoff projection | The payoff block of the loan summary, computed on each read for a good-through date. It is not a resource: nothing is stored, and there is no quote ID or quote lifecycle. |
| Form entry | A typed answer set captured against one customer or loan under a published form definition, governed by permissions derived from the form key. |
| Reconciliation exception | A tracked difference between provider events, money intents, and the ledger — stuck-before-submit, settled-not-posted, and similar — worked through the recon-exceptions routes. |
| Document / Note | Lending Core records linked to customers and loans, with scanning, history, revisions, and redaction. Distinct from servicing case documents and case evidence under /v1/servicing. |
LendEasy-Tenant |
The required header naming the tenant on every API call. |
LendEasy-Case |
The header carrying an open case reference on human servicing mutations when case management is enabled; see Access control. |
Foundational terms
A few terms describe how the platform itself is put together. For the full picture, see System architecture.
| Term | Meaning |
|---|---|
| Control plane | An operating layer that holds a desired state, observes authoritative state, reconciles the difference, governs execution, and records every decision. |
| System of record | The lending core that owns loans, balances, schedules, payments, and accounting. On the full platform that is the LendEasy Lending Core; with bring-your-own-core, LendEasy references your system and never becomes the financial record itself. |
| Restriction | A typed hold — payment, autopay, disbursement, collections, communication, or SCRA/MLA protection — applied at customer scope, optionally to one loan, and enforced at execution time until cleared. |
| Drift | A divergence between LendEasy’s operational state and the authoritative state in your core, surfaced as work rather than hidden. |
A missing fact is never permission. Where a rule needs a fact the platform does not have, the outcome is missing facts and the work routes for review — it never falls back to a default that allows the action.