Use case · Collections and arrears
Agent Serve drafts each day’s arrears letters from approved templates, flags wording to avoid, routes vulnerable customers to specialists, and waits for approval.
A collections and arrears use case for Agent Serve, the AI workspace where teams build, deploy, and manage AI agents: a scheduled workflow reads accounts in arrears from a table, drafts letters from approved templates, checks wording with Guardrails, routes flagged customers to specialists with rules and waits for a collections lead to approve.
At a glance
The problem it removes
Every morning the team exports the arrears list, picks a letter template for each account, and edits it by hand. Wording drifts, customers who need extra care are spotted late, and when a regulator asks how a customer was treated, the answer is spread across inboxes and notes.
How it runs
Agent Serve drafts and checks every letter. A collections lead approves what customers receive.
A schedule starts the workflow every weekday morning, and the Table block reads accounts at each arrears stage.
A rule flags accounts with a vulnerability marker or a recent hardship note, and those go to a specialist instead.
An agent drafts each letter from the approved template for its stage, and Guardrails flag any wording on your avoid list.
A Human step sends the batch to the collections lead, who approves, edits or holds each letter before it goes out.
Before and after
| Work | Before | With Agent Serve |
|---|---|---|
| Letters | BeforeTemplates copied and edited by hand, account by account. | With Agent ServeDrafted from the approved template for each arrears stage. |
| Wording | BeforePhrases to avoid are caught only in sample checks. | With Agent ServeGuardrails flag them in every draft before review. |
| Customers who need care | BeforeHardship notes are found after standard letters went out. | With Agent ServeA rule sends flagged customers to a specialist first. |
| Evidence of treatment | BeforeSpread across inboxes, call notes and spreadsheets. | With Agent ServeEach contact and approval is recorded with the run. |
In depth
Agent Serve gives collections leads the drafts, the checks and the approvals in one place.
An Agent Serve Relay drafts a payment arrangement confirmation and stops. The lead sees the exact email and approves or declines it.
An Agent Serve rule flags accounts with a vulnerability marker or a hardship note, and the workflow sends them to a specialist queue.
The Agent Serve Permissions step limits a Relay’s email to internal people, an allowlist or anyone you allow, and names who approves.
One step of the process: the Agent Serve Call Compliance dashboard scores uploaded call transcripts against your checks, with a pass or fail per call.
What does the work
An arrears table holds each account’s stage, last contact and notes.
A scheduled workflow drafts letters and checks their wording.
Rules send customers who need extra care to a specialist.
A Relay drafts arrangement confirmations and waits for approval.
A dashboard shows contacts, arrangements and broken promises by stage.
Examples
Three collections processes teams run on Agent Serve, each with a person approving what customers receive.
An Agent Serve schedule drafts each morning’s letters from the approved templates, and a lead approves the batch.
An Agent Serve Relay watches the arrangements table and drafts a confirmation for each new row, sent after approval.
An Agent Serve rule flags accounts with a hardship note, and the workflow moves them to a specialist queue before any letter.
Examples show what teams can build in Agent Serve. They describe patterns, not named customer deployments.
Governance
Agent Serve keeps a person on every customer message and records what each run did.
Security and governanceAn Agent Serve Relay can be limited to internal recipients, an allowlist or anyone, and never sends email without approval.
The Agent Serve Guardrails block checks each draft against patterns you set, such as phrases to avoid, before it moves on.
Relays never send email or book meetings without an approval, under any autonomy setting. An approval covers the exact payload the reviewer saw and expires if nobody decides.
Proposals, approvals, declines and executed actions are written to the workspace audit log, which admins can export.
Every workflow run records each block’s inputs, outputs and errors, so a reviewer can trace how a result was reached.
Admin, write and read roles per workspace, plus permission groups that limit which models, blocks and integrations a team can use.
Sign in through your identity provider with SAML 2.0 or OIDC, and provision people with SCIM.
Run Agent Serve in your own cloud or data centre, next to the systems it reads.
Zero data retention with supported model providers, backed by your own provider keys and agreements or self-hosted models. Some provider features keep data by design, such as Gemini Deep Research, so they are outside zero data retention.
A Guardrails block detects personal data and blocks or masks it before text reaches the next step.
Further reading
Collections is where conduct rules meet customers under pressure. This guide compares four ways to check calls and letters, what each one catches, and how they combine in Agent Serve.
Consumer lending · 7 min readCustomers often mention a bereavement, an illness or a lost job in passing, and the call moves on. Checking every call can catch those moments. What happens next has to stay with a person.
Retail financial services · 6 min readDisputes that fail reconciliation tend to sit in a side queue until the chargeback window has closed. Here is an end-to-end setup that finds them early, ranks them by deadline and drafts the pack, with an analyst signing every filing.
Cards and payments operations · 12 min readQuestions
No. In this Agent Serve process a collections lead approves every letter, and a Relay never sends email without approval of the exact message.
No. An Agent Serve rule only routes accounts that already carry a marker or a hardship note to a specialist, who decides what happens next.
Only as one step. Call checks live in Agent Serve Dashboards, where the Call Compliance template scores uploaded collections call transcripts against your checks.
Agent Serve data sources sync it into a workspace table, or a workflow can read it from your database or loan system through an integration or API block.
A team can start with the Agent Serve arrears table and the morning letter workflow with wording checks, then add specialist routing and a Relay for confirmations.
Last updated
Enterprise access
Agent Serve is available to teams by invitation. Tell us what you want to build and we will set up your workspace with you.