How RonCredit is scaling loan support without scaling its contact centre
EU
First deployment in a regulated industry
GDPR and EU AI Act compliant
-70%
Target reduction in contact centre load
Business growth without headcount growth
1 month
From kick-off to production
Including proprietary CRM integration
Romanian microfinance, built for speed.
RON Credit IFN SA is a Romanian non-banking financial institution, registered with the National Bank of Romania. The company issues consumer microloans entirely online: 100 to 10,000 lei, for terms of 3 to 30 days, with the option to extend. An application takes 10 minutes, and funds reach the customer's bank card within minutes of approval.
RonCredit launched only a few months ago and brought Obriy AI in as its AI provider almost immediately, so support automation is built into the operating model from day one, not bolted onto an already oversized contact centre.
RonCredit
Growth that scales the contact centre one-for-one.
Financial services in the EU are regulated, so any automation of borrower communication has to comply with GDPR and the EU AI Act, and must never give the customer financial advice or make commitments on the company's behalf.
RonCredit's strategic goal is simple: grow. But in online lending, growth in loan volume automatically drives growth in support tickets, and the conventional answer is to hire more agents. For a young company with eight agents, that means the cost of support scales with revenue instead of independently of it.
Peaks make it harder. Contact centre load is uneven and tied to marketing activity: the moment a campaign goes live, ticket volume jumps. Staffing for peak load makes no economic sense, and not staffing for it means service quality drops exactly when the company is paying for traffic.
An AI workflow that acts directly in the CRM.
Obriy AI deployed an AI workflow for RonCredit on the SURE platform, integrated deeply via API into the client's proprietary CRM. The agent covers borrower support and onboarding: it receives a ticket, categorises it, looks for an answer in the knowledge base, and either responds itself or escalates to a human agent, performing its actions directly in the CRM.
Categorisation
The system receives the enquiry and identifies the topic: how to get a loan, application requirements, transfer methods, amount and due date, extending the term, reasons for rejection, early repayment.
Knowledge lookup
The agent queries the RonCredit knowledge base and checks whether it holds enough information to answer this topic.
Answer or escalation
If the knowledge is there, the borrower gets an automated answer. If not, the enquiry is escalated to a human agent with full context preserved.
Actions in the CRM
The agent performs operations directly in the client's CRM via API: replying in the conversation, changing statuses, handing the enquiry to human agents.
Deep integration into a proprietary CRM.
Executing actions, not just relaying messages
RonCredit's communication runs on a CRM built specifically for the company. Obriy AI integrated into that system deeply via API, executing actions inside it rather than simply relaying messages. At this stage the integration covers the communication tooling; the next step is access to customer data, so the agent can answer from the context of a specific loan and not only from the knowledge base.
The knowledge base itself is built inside the SURE platform. Updating knowledge requires no changes to the CRM and no releases on the RonCredit side: the team works with the base in SURE, and the agent starts answering under the new rules.
Compliance as part of the architecture, not a bottleneck.
The solution complies with GDPR and the EU AI Act, sufficient to operate in the regulated EU financial services industry. The agent carries a built-in disclaimer that it does not provide financial advice, and topics that require a human decision are deliberately kept out of automation and escalated to agents.
Sensitive topics don't get automated by accident. They're kept out on purpose, and escalated to a person every time.
RonCredit
Live in production. Deployment took roughly one month, kick-off to go-live, including integration into a proprietary CRM.
Because the system launched recently, final automation percentages are still taking shape. The target agreed with the client is a 70% reduction in contact centre load.
-70%
Target load reduction
On the contact centre
1 month
From kick-off to production
Including proprietary CRM integration
8
Agents serving a growing base
No hiring for peaks
What this gives the business
- Growth in loan volume no longer means proportional growth of the contact centre: 8 agents serve a larger customer base than they could without the AI agent.
- Marketing peaks stop being an operational risk: the agent absorbs the spike in enquiries without hiring people for a campaign.
- Compliance is not a bottleneck: GDPR and the EU AI Act are accounted for in the architecture, the financial-advice disclaimer is built in, and sensitive topics are escalated to a person.
- Borrowers get answers to routine questions immediately, at any hour, not only during contact centre working hours.
Roadmap.
Proactive scenarios
The agent not only answers but initiates contact at the right point in the customer journey.
Deeper CRM integration
Access to customer data so answers are grounded in the context of a specific loan, not only the knowledge base.
Voice channel
Added alongside WhatsApp and web chat.
Beyond support
Extending the system to other repetitive business processes inside the company.
- Platform
- SURE by Obriy AI
- Architecture
- AI enquiry-handling workflow: categorisation, knowledge lookup, escalation
- Channels
- WhatsApp, website chat; voice planned
- Integration
- Deep API integration into RonCredit's proprietary CRM, with action execution
- Knowledge source
- Knowledge base built inside SURE
- Compliance
- GDPR, EU AI Act; built-in disclaimer on not providing financial advice
- Languages
- Romanian, English
- Deployment time
- Around one month, kick-off to production
- Deployment
- Fully delivered by the Obriy AI team
Ready to grow without growing your contact centre?
See how Security → keeps regulated deployments compliant by design.