AI collection engine · NPL & legal

The engine behind the AI‑first servicer.

Vexor reads every court document, knows every past call and decides the next move on each case. Then it does the work, inside your CRM. Your people keep the judgment calls.

CASE 0482 · UNSECURED · €3,240ILLUSTRATIVE
COURT
Enforcement title granted, 2 Sep
ASSETS
No attachable income on file today
CALLS
14 Sep: “I start a new job in November.”
PAYMENTS
Plan broken in June after 2 instalments
DECISION
Willing, not able yet. Call on 3 Nov at 13:00 with a plan sized to the new salary.
Why: the debtor gave a date, and the first salary lands end of October. No call before then.
Next action and notes written back to the CRM
IN PRODUCTIONWhitestar, €10bn+ AUM
PILOT TO PRODUCTION3 months
COMMERCIAL MODELPaid on outcome
DEPLOYMENTInside your own tenant
The old model

To grow 2×, a servicer hires 2× people.

Every case runs on people reading documents, reviewing files and making calls. The market shrinks, but the back office stays the same size, and small tickets cost more to work than they return.

The super agent

Not a phone bot. Not a document reader. One engine that runs the whole case.

A person works 30 cases a day and skims the file. Vexor reads every case, every morning, with the full context.

  1. 01

    Understand

    Court documents, asset reports, payment history and every past call, in one case file.

  2. 02

    Decide

    Who to pursue, when, and with what. If nothing has changed on a case, nobody calls it.

  3. 03

    Act

    It calls and negotiates within your limits, files the legal step in your CRM, or waits for the right date.

  4. 04

    Learn

    Every outcome updates the case, and every case makes the engine better.

Your whole book, sorted every day.

ILLUSTRATIVE SPLIT
  • UnreachableFind a working number or park it.
  • Willing and ableReach them fast and close.
  • Willing, not yet ableBook the date they gave and call then.
  • Able, not willingBring the facts, the terms and the legal position.
  • BlockedGet the missing document, then resume.
  • StopNo more contact.
In production

Whitestar runs part of its core collection on Vexor.

Read the case study (PDF) →

“AI has allowed us to enter segments that we previously discarded due to negative economics. This unlocks a new line of revenue that supports our growth even in a shrinking NPL market.”

WhitestarMarco Freire, CEO
LEGAL20%of the cost for the work of a 5-person legal team
LEGAL95%accuracy. Suspected errors go to the team for review
VOICEDailycollections since August, on first- and third-party portfolios
TIME TO VALUE3 mofrom pilot to production, built directly into their systems
How it is deployed

Built to pass IT, compliance and the CFO.

Runs in your tenant

Deployed on your cloud with your approved models. Your data never leaves your infrastructure.

Works in your systems

Sits on top of your court feed, CRM and dialer. Your team keeps working the way it does today.

You set the rules

You approve every script and every payment limit. The AI checks its own work, and edge cases go to your people.

Paid on outcome

You work on success, and so do we. Cost only rises when there is output.

Servicer 2.0

More cases. Better quality. Structurally lower cost.

One engine across every market and portfolio. What it learns on one book makes the next one better.

Alen TadevosyanFounder & CEO