Short answer: In a collection agency or a bank's or microfinance lender's recovery team, results depend on calls, and so does risk: one rude or threatening call can become a complaint and a reputational problem. dooCall automatically records calls on collectors' and field agents' ordinary SIM phones, checks them with AI against your collection script, and shows where a field agent was at the moment of each call.
Why is the call the key moment in collection?
Most contact with a debtor happens by phone, and that's where agreements are made. SMS and letters remind, but a payment date, instalments or restructuring are usually agreed in a live conversation.
Typical problems in recovery teams:
- A collector dials hundreds of numbers a day, but the manager sees only the summary report, not call quality.
- The debtor says "I was told I could pay later" and there is no way to check.
- Some staff slip into pressure, threats or insults: complaints, inspections and reputational risk.
- The debt gets discussed with third parties (relatives, the workplace).
- A field agent says "I went to the address" but nothing confirms it.
- The debtor doesn't answer an unfamiliar number.
Hiya's 2024 report found 77% of people are more likely to answer if they know who is calling. That is why it matters to call from a stable, recognisable company number.
How does dooCall work in a recovery team?
dooCall is installed on the Android phones of collectors and field agents and records every incoming and outgoing call on the company SIM. Recordings are stored encrypted on servers in Uzbekistan.
Managers see in the cabinet:
- Daily calls, answered calls and talk time per employee.
- Full call history by debtor number.
- Unanswered outbound and missed inbound calls.
- Where a field agent was at the moment of each call.
- Reports by period, employee and debtor, with export.
With Odoo, Bitrix24 or amoCRM connected, the recording lands on the debtor's record.
What does AI analysis check in a collection call?
AI analysis is live, and Deep Vision configures it to your collection script and internal rules. For example, AI can check in every call:
- Did the collector introduce themselves and the company, and say the call is recorded?
- Did they verify identity: is this really the debtor?
- Were the amount, due date and consequences stated correctly and clearly?
- Was a specific payment date and amount agreed?
- Were there threats, insults, humiliation or unlawful pressure?
- Was the debt discussed with a third party?
Managers don't listen to thousands of recordings, only the risky or unsuccessful calls AI flags. AI analysis costs 549 so'm per minute and can be applied to selected calls rather than all of them.
Call-time location for field agents
dooCall logs where the employee was when a call was made or received; if there is no call, no location is sent. It is not continuous GPS tracking.
The practical rule for field collection: on arrival at the debtor's address, the agent makes a short call to the debtor or the office. That call shows the visit time and address in the cabinet, and the conversation itself is recorded. It helps verify reports like "I went, nobody was home".
What changes with dooCall?
- Promises are logged. "I'll pay by Monday" stays on record and the next call builds on it.
- Conduct standards hold. Threats and insults don't stay hidden; AI flags them.
- Complaints are reviewed on facts. If a debtor complains, you can listen to the call and see who is right.
- Staff work is measured. You see how many debtors each employee reached per day and how many agreements they made.
- Field visits are confirmed. Call-time location shows the visit time and address.
- The number stays with the company. If an employee leaves, the history of debtor conversations isn't lost.
How should a collection call be structured?
A good collection call isn't pressure; it's a conversation that ends in a specific agreement.
- Introduce yourself and the company, and say the call is recorded.
- Verify identity; if it isn't the debtor, don't mention the debt.
- State the amount, due date and overdue status calmly and clearly.
- Ask about the reason for the delay and listen.
- Agree a specific date and amount, or offer an instalment plan.
- Confirm the agreement in writing by SMS or Telegram.
The same points become the checklist for AI.
The legal side
Debt calls involve personal and financial data. Uzbekistan's Law ZRU-547 "On Personal Data" applies, and since 16 April 2021 citizens' personal data must be stored in Uzbekistan; dooCall recordings are stored on servers in Uzbekistan. Get written consent from employees, tell debtors calls are recorded, never discuss the debt with third parties, and allow no threats or insults. Banks and microfinance lenders should also consider their sector requirements. Consult a lawyer; this is not legal advice.
Frequently asked questions
Does dooCall dial automatically? No. dooCall records the calls employees make and receive themselves.
Is location tracked all the time? No. Location is sent only at the moment of a call. Without a call, the employee's whereabouts are not logged.
Does it work on iPhone? No. dooCall runs on Android 10 or later; there is no iPhone version because iOS does not allow call recording.
How much does it cost? 49,000 so'm per phone per month, with a 14-day free trial. AI analysis is separate, at 549 so'm per minute.
Can a recording be used in court? That's a question for a lawyer. Get legal advice before using a recording.
Sources
- Hiya, 2024 State of the Call: https://blog.hiya.com/2024-state-of-the-call-consumers-prefer-voice-but-spam-and-fraud-are-threats
- Law "On Personal Data" (ZRU-547): https://lex.uz/docs/4831939
- DLA Piper, Uzbekistan data protection law: https://www.dlapiperdataprotection.com/index.html?t=law&c=UZ
- dooCall official site: https://doocall.uz
Related: Instalment sales and payment reminder calls · Is call recording legal? · AI call analysis
To set up dooCall and AI script checks for your recovery team, send a request or call +998 77 093 00 07.