Skip to main content

AI

September 24, 2025

 

From risk to resilience: How AI is helping entrepreneurs unlock credit in Colombia

Named for Inca system of knots used for record-keeping, Quipu uses artificial intelligence to untangle funding for Latin America’s informal businesses.

Quipu, led by Mercedes Bidart, right, with one of the entrepreneurs she serves, uses AI to help small businesses show they are creditworthy and access loans they need to grow. (Photo courtesy of Quipu)

Aimee Levitt

Contributor

When Rosa started sewing brightly colored handwoven totes and bucket bags at her home in Barranquilla, Colombia, she was a one-woman operation, enlisting her daughter to help promote her creations on Instagram and Facebook.

As her eponymous label caught on, she began looking for ways to hire staff to help keep pace with demand for her brand and eventually open a retail space in her Caribbean coastal city. That’s when she hit a roadblock: Despite the steady growth of her business, she struggled to get the financing she needed.

Like many Latin American workers, Rosa is part of what’s known as the informal economy. Though they comprise a third of Latin America’s economy, most operate without bank accounts, using cash or digital wallets. This can be fine for ordinary daily transactions, but not when it comes to applying for a formal bank loan. As in the U.S., Latin American banks require proof of a substantial financial and credit history. And without that information, entrepreneurs like Galván cannot reach their full potential — much less bridge the substantial wealth gap in Latin America.

“When you’re born poor,” explains Mercedes Bidart, “it’s hard to build credit.”

As co-founder and CEO of the financing platform Quipu, Bidart wants to change that. Quipu is named after the ancient system of knots the Incas used for record-keeping, but its superpower is state-of-the-art artificial intelligence. It uses AI to sort through nontraditional data, including a company’s social media, inventory and real-time financial transactions, to determine whether businesses like Galván’s is creditworthy, putting them on the path to securing funding.

AI holds immense potential to deepen impact and scale essential services. That’s why the Mastercard Center for Inclusive Growth and data.org launched the Artificial Intelligence to Accelerate Inclusion (AI2AI) Challenge last year, to help startups, social enterprises and nonprofits develop and scale AI solutions that advance inclusion and economic empowerment.

“AI holds extraordinary potential, but its true power is realized only when everyone can access and benefit from it,” says Uyi Stewart, the Center’s vice president for Inclusive Innovation and Analytics. “By breaking down digital and informational divides for underserved communities, we unlock a future where innovation is inclusive, economic growth is amplified, and opportunities are shared by all.”

A jewelry entrepreneur clasps a bracelet on Mercedes Bidart's wrist.

 

Quipu was one of five winners of the challenge and is using the $200,000 prize money, along with technical assistance and mentoring, to add a WhatsApp tool called Edubot that supports businesses’ daily needs — Bidart describes it as “a CFO in your pocket.”

Since its founding in 2021, Quipu has allocated $5 million and helped more than 26,000 Colombian entrepreneurs, 50% of whom are women. Its loans range from $100 to $5,000; the average is $300. Now Quipu is licensing its technology to other lenders in Colombia with an eye toward expanding throughout Latin America.

Bidart conceived of Quipu as part of her master’s thesis in city planning at the Massachusetts Institute of Technology. She arrived at the prestigious university in 2017 fresh from a job in a public policy think tank organization in her hometown of Buenos Aires, where she became convinced that technology — with its lack of borders — could put money in the hands of unbanked people far more efficiently than social policy.

To assess creditworthiness, we’re not basing our decisions on what you paid in the past. We want to understand your present and your potential.

Mercedes Bidart

 

At MIT, she began studying ways to determine creditworthiness without relying on traditional financial documentation; during her first semester, she won a grant to support her research. At a conference in Boston, she met with the founders of a company in Colombia, who invited her to work with informal businesses there. Using those contacts as her first clients, Bidart gradually developed Quipu, putting her convictions and unconventional financing methods into practice.

“To assess creditworthiness, we’re not basing our decisions on what you paid in the past,” Bidart says. “We want to understand your present and your potential.”

Quipu relies on non-traditional data because that information can provide a clear picture of a company’s present and potential, not just its past. The company’s algorithm has amassed enough data over the years so that it can determine whether a business is credit-worthy within a few seconds.

“I really think loan offers should be personalized,” Bidart says. “Everything is really standard right now. We want to offer help with what the user really needs.”

While Quipu initially underwrote the loans itself, Bidart is now looking for financial partners that can use the scoring technology to allow more entrepreneurs to access capital to grow ventures that support their families and communities.

“Those loans are coming at a really important moment,” Bidart says. “After a business gets a loan, it increases its income by 9%.”

That fortuitous timing has certainly worked out for Galván. Today, with the help of six Quipu loans, she’s opened the retail space of her dreams, as well as a workshop that employs half a dozen employees.

“She’s been able to grow,” Bidart says, “because she has capital.”

Using data to think bigger for small businesses

When put to work intelligently, data could be the secret ingredient to giving small business owners needed boosts, writes Mastercard's Jess Turner in Forbes. 

A man in a plaid shirt in a workshop looks at a computer.