Office & managementNewVerified Real deployment at a named or precisely described company, with a measured result from a primary source.

AI collections: prioritised worklists and automatic reminders get customers to pay on time (EZG Manufacturing, Yaskawa, Lodge Cast Iron)

Added 2026-10-21 · Last updated 2026-10-21

Role

Finance (accounts receivable: credit and collections) · Typical users: AR clerk, collections specialist, credit manager, controller, CFO of a small or mid-size manufacturer

The problem

In many plants, collections is a part-time job: the same one or two people also handle cash application, billing and other finance work. Reminders go out when someone has time, the worklist is a spreadsheet or the aging report, and the biggest or most at-risk accounts are not always called first. Past dues grow, days sales outstanding (DSO) creeps up, and the company finances its customers. Lodge Cast Iron ran collections and dunning in spreadsheets with faxes and Outlook; Yaskawa America's past-due receivables were rising because the collections worklist was prioritised by hand; EZG Manufacturing's two-person AR team had a clear strategy but could not execute it consistently next to its other duties.

How AI helps

The collections software connects to the ERP's open invoices and payment history. It ranks accounts by amount, age and the customer's payment behaviour so the team works the right ones first, and it runs the routine outreach on its own: due-date reminders, follow-up emails, automated phone calls at EZG, with each reply and promise to pay logged. Disputes, short payments and sensitive accounts go to a person. The AR team decides on holds, payment plans and escalation.

Step by step

1. Record the baseline: DSO, past dues by aging bucket, hours a week on collections, and the ten customers that pay latest.
2. Write the rules down: reminder schedule before and after the due date, tone by customer segment, who may offer a payment plan or put an account on credit hold.
3. Clean contact data: the right accounts-payable contact and email for each customer, and the portal or remittance address they use.
4. Start with due-date reminders only, as EZG did, then add follow-up emails and calls once replies are handled well.
5. Route disputes, short payments and key accounts to a person with the invoice, delivery proof and history attached.
6. Review monthly: DSO, past dues, promises kept versus broken, and complaints; adjust the sequence by customer segment.

Tools

• Stuut (AI collections outreach by email and phone): used at EZG Manufacturing
• HighRadius (AI-prioritised collections worklists, automated correspondence, deductions): used at Yaskawa America and Lodge Cast Iron
• Other tools to evaluate: Esker, Billtrust, Versapay, Upflow, Chaser, Tesorio, and the collections module of your ERP (SAP, Microsoft Dynamics 365, NetSuite)
• Any LLM (ChatGPT, Claude, Copilot, Gemini) with the prompt below, on your aging report, to build this week's call list and draft reminders

As reported by the source; most figures are vendor or self-reported and not independently verified.

Human review: cautions

• Customer relationships come first: disputes, key accounts and sensitive cases stay with a person, and sales should know before an important customer is chased hard.
• Automated reminders only work with accurate invoices and contacts; fix billing errors and missing delivery proofs, or the system chases the wrong debts.
• Credit holds and payment plans remain decisions for the credit manager.
• All figures are vendor-published; the Lodge Cast Iron figures come from a summary of the vendor's case.

You are an experienced collections specialist at a manufacturer.

I will paste: (1) our open receivables [customer, invoice number, invoice date, due date, amount, currency, days past due, any dispute or note]; (2) payment history for the last 12 months [customer, invoice, due date, paid date] or average days late per customer; (3) our collection rules [reminder schedule, credit limits, who approves holds or payment plans, customers to handle personally]; (4) last contacts and promises to pay, if any.

Do this:
1. Build this week's worklist as a table ranked by priority: customer, total past due as listed, oldest invoice, usual payment behaviour, reason for the rank. Copy amounts as given; do not add up or convert currencies yourself, and give me a spreadsheet formula for any total I need.
2. For each customer, propose the next action (friendly reminder, firm reminder, phone call, escalate to sales or credit manager) following our rules.
3. Draft the reminder emails in a polite, firm tone, quoting invoice numbers, amounts and due dates exactly as pasted.
4. List accounts with disputes or notes separately, with the question to resolve first.
5. Flag customers close to or over their credit limit, for the credit manager to decide.

Never threaten legal action or a credit hold unless our rules say so.
Use only the data I paste in; if something is missing, write 'not found' or list it as a question instead of guessing. Do not send, order, post or change anything in any system: produce drafts for my review and approval.

Related cases

• AI cash application and deduction management: read complex remittances, track customer deductions
• AI answers customers' 'where is my order?' emails from live ERP data (promotional-products maker, medical device maker)
• AI accounts payable invoice processing
• An AI agent that turns your emails and messages into a live, prioritised to-do list and day plan

All 9 Office & management use cases →

Sources

• https://www.stuut.ai/case-studies/ezg-manufacturing
• https://www.highradius.com/en-gb/resources/case-studies/yaskawa/
• https://www.casestudies.com/company/highradius/case-study/how-lodge-cast-iron-fueled-business-growth-13-reduction-in-past-dues-5-day-increase-in-dso-efficiency-and-96-net-recovery