How R2 cut its collections operation costs by 70% and scaled it to 5 countries with AI agents

Scaling a collections operation usually means one of two things: hiring more people or hiring traditional collections agencies with limited capacity. For R2, one of Latin America's most prominent fintechs, neither was an option. With a portfolio growing fast across several markets in the region, R2 wanted to grow without inflating the team and without losing the efficiency they're known for.

A perfect use case for AI agents.

Who R2 is

R2 is the leading embedded lending platform in Latin America. Founded in 2020 by Roger Larach and Roger Teran and backed by Y Combinator, it is headquartered in Mexico City and enables digital platforms such as payment processors, marketplaces, point-of-sale systems, and mobility apps to offer financing, via API, and without the platform taking on credit risk, to the participants who generate revenue within them: from merchants and restaurants to drivers and gig-economy couriers.

Its revenue-based financing model uses machine-learning underwriting on transactional data (not traditional credit history), which lets it extend credit to merchants and independent workers that banks tend to leave out. A good example is inDrive Money, where R2 lends to drivers and repayments are automatically deducted from their per-trip earnings. In a region where the vast majority of SME credit demand goes unmet and the financing gap exceeds one trillion dollars, R2 tackles a problem of continental scale.

Its track record speaks for itself:

• More than 100,000 SMEs served and thousands of independent workers (drivers and couriers) through partners like Uber Eats and inDrive across five markets: Mexico, Chile, Colombia, Peru, and Brazil.
• Partnerships with the region's most recognized platforms: Uber Eats, Rappi, Clip, PayU, inDrive, among others.
• Backing from world-class investors: Gradient Ventures (Google's AI fund), General Catalyst, Y Combinator, Endeavor Catalyst, Cometa, Hi Ventures, and FEMSA Ventures, among others.
• In October 2025, Ant International (the global arm of Ant Group, co-founded by Jack Ma) made a strategic investment in R2, adding world-class risk and AI capabilities.

The challenge: scaling the operation without becoming a call center

Before Altur, R2's collections operation relied on a mix of external agencies and limited internal efforts. Internally they used email, SMS, and some WhatsApp interactions, but without the capacity to cover 100% of the base. That forced them to constantly prioritize which customers to contact and to leave part of the portfolio out.

The real bottleneck was scalability. The volume of accounts grew month over month, but the model based on people and agencies couldn't scale at the same pace. R2 needed an automated solution to sustain growth without turning collections into a heavy, costly operation.

Why they chose Altur

After reviewing different providers, the factors that made the decision to try Altur easy (in their words) were:

• High-quality natural interactions, calls with a truly human-like AI voice and automatic, granular evaluations on every call. The agents negotiate, offer alternatives, and close agreements just as the best human operator would, on every call.
• Multichannel, the agent can move the conversation to WhatsApp without losing the call's context and keeping the interaction personalized, across thousands of interactions at the same time.
• Speed, in the words of Angel Coronado, Director of Collections at R2:

"We began to see results from the very first month of implementation. From that point, an improvement in operational efficiency and in the performance of key metrics was evident."

Unlimited calling capacity

The main factor behind the success of R2's collections operation with Altur is, without a doubt, unlimited contact capacity: with Altur's AI agents, contact capacity is no longer tied to team size. Altur can sustain thousands of calls in parallel, so it's no longer necessary to prioritize the base due to lack of staff: 100% of the portfolio can be covered consistently, at the right time and through the right channel.

The combination of capacity and precision in every contact results in an extremely efficient operation. Scalability and savings are built on top of it.

Rapid expansion across five Latin American countries

R2 now runs its collections with Altur in five countries in Latin American, and keeps expanding. What made this possible was moving into new markets without scaling the collections team at the same pace.

Each country gets a localized experience: local phone numbers and agents with a local accent, which meaningfully improves contact rates. Instead of copying and pasting one operation across markets, every customer hears something that feels local to them.

70% reduction in operating costs

The numbers speak for themselves:

• 70% reduction in operating costs compared to the previous model of people and external agencies.
• Across 8 different products, Altur's AI agents matched the results of the human operation and even outperformed it by up to 15%.
• A significantly higher ROI than external agencies.

The most important improvement for their operation, according to R2 itself, was being able to scale efficiently across multiple countries without needing to grow the team.

Do they recommend it?

We asked Angel whether he'd recommend Altur to other financial institutions. His answer was a straightforward yes.

If your collections are limited by the size of your team or the agencies you rely on, or you want to expand into new markets without growing costs, let's talk. Altur helps fintechs and banks collect more effectively, across more countries, and at a fraction of the cost.