David Vélez on Nubank, Financial Access and AI Banking

In this September 2023 interview, Nubank founder and CEO David Vélez tells Harry Stebbings about interviewing at Sequoia. Doug Leone met him first and asked about his family and character rather than spending most of the time on his résumé. Vélez saw the senior attention and rapid follow-up as evidence of the firm’s culture.

While in business school, he helped explore Sequoia’s Latin American investments. He recalls early calls to Brazil, full school days and short trips with Leone to meet companies in São Paulo. The pace was intense, but he found the experience stimulating.

Vélez credits Leone with an unusual ability to read people. He says he learned to look for character and motivation and to value people with exceptional strengths, even if they also have weaknesses. These are his hiring preferences, not validated personality tests.

At Sequoia he was asked for his opinion in investment meetings even as an intern. He brought that sense of partnership to Nubank, he says, seeking a culture in which people at different levels prepare, speak up and feel responsible for decisions.

In 2012, after he had spent months preparing a Latin American operation, Leone told him Sequoia would not open the office. Vélez says the decision was a shock but he understood the firm’s concern about the opportunities it had seen. He could move to California or pursue a long-held ambition to start a business.

Stebbings questions whether Latin America has enough venture-scale exits for growth investors. Vélez disagrees, arguing that the region’s large population and unmet needs provide opportunity. His comparisons of regional markets and his forecast of more large technology companies are his views, not independently verified market analysis.

He says that in the early 2010s many entrepreneurs were copying US ideas rather than addressing local needs such as financial access. Founding Nubank meant challenging conventional expectations that established banks would overwhelm a new entrant. He believes attitudes and capital availability have since changed.

Vélez points to local public markets, US listings and acquisitions as possible sources of liquidity, while acknowledging that smaller individual countries can be harder markets for exits. His examples illustrate possible routes; they do not guarantee liquidity or venture returns.

Asked why Nubank has “won,” he resists that conclusion and says it has much left to prove. He reports that close to half of Brazil’s adult population was a customer around the time of the interview. This is a 2023 company claim, not a current population estimate.

He says different customers need different products: high-income users, older adults and children may each find gaps or complexity in the existing offering. The company’s aspiration is increasingly personalized service, but he describes work still to be done rather than a completed capability.

Nubank chose to become customers’ primary bank rather than only an ancillary wallet. Vélez says obtaining banking licenses entails both advantages and obligations and makes expansion more local. The company therefore chose to go deep in Brazil, Mexico and Colombia rather than enter many markets rapidly.

He says roughly 60% of its Brazilian customers used Nubank as their primary account, implying close to 30% of Brazilian adults under his stated customer-base estimate. These are company-reported figures and definitions at that time, not independently verified current market shares.

Vélez sees complacency as the company’s greatest threat. He cites differing shares across credit cards, personal loans, investments and insurance to argue there is room to grow, and describes plans for a broader consumer platform. The market-share figures and growth prospects are his account and forecast.

To maintain urgency as the company grows, he looks for people motivated by building rather than a brand on their résumé. He recalls an early São Paulo house with 20 people and limited furniture, which he says attracted some candidates and deterred others. Now the company relies more on interviews and culture.

He says Nubank celebrates briefly before asking what comes next, and acknowledges that some employees may want more celebration. He treats that tension as part of maintaining urgency. This is his leadership approach, not a tested formula for every organization.

Vélez believes a bank can expand beyond financial services because a primary banking relationship can build trust, a customer base and useful data. He contrasts this with commerce or social platforms adding finance. His claim that the path is easier for banks is a strategic hypothesis, not an established outcome.

He claims Nubank’s Mexican credit card had a net promoter score of 94 and calls it the highest-rated consumer product globally. The interview provides no independent comparison or survey methodology, so the global ranking is unverified. He has a direct commercial interest as Nubank’s CEO.

He says Nubank’s marketplace connects customers to more than 180 partners offering products and services, with credit as one component. The company acts as a platform rather than running the merchants’ logistics. Customer counts, activity and the claim of superior credit pricing are company statements, not independently established facts.

Asked why allocate resources to a marketplace when financial products remain underpenetrated, Vélez says roughly 80% of the company’s resources still went to financial services. He sees the marketplace as a longer-term growth and diversification opportunity. Its eventual resilience and returns were not established.

Vélez argues that Brazil, India and China have made progress in digital payments where financial exclusion created pressure for change. He criticizes aspects of US and European regulation as less welcoming to fintech. These are his business and policy judgments, not a comprehensive comparison of regulatory systems.

His advice to founders is to consider the obligations and benefits of becoming a regulated bank rather than avoid a license solely to seek a higher valuation. He also wants regulators to recognize gaps in access. Licensing decisions and legal obligations depend on jurisdiction and business model.

Asked whether digital banks could serve the wealthiest clients, Vélez says it may begin with simpler products before extending further. He predicts that AI could eventually offer sophisticated advice at lower cost. That is a forecast by a bank CEO; investment advice can involve conflicts, errors and individual needs regardless of the delivery method.

He argues that smartphones lowered the cost of reaching customers without branches and made smaller accounts and loans more feasible. His cost comparison is a company claim in the interview. He sees financial literacy and complexity as remaining barriers to broader access.

Vélez imagines AI as a “banker in every pocket” that helps customers navigate decisions. The claim that it could democratize advice is a vision, not evidence of accuracy, suitability or equal access. Nubank has a direct interest in developing and selling financial products.

He says the company had not seen a need to reorganize around AI and instead wanted to embed it across existing teams. He hoped AI could improve products or reduce costs, but the discussion provides no measured savings or demonstrated outcomes.

The team was debating whether an AI banker should push Nubank products or maintain enough neutrality to recommend another provider when appropriate. Stebbings notes that transaction fees and incentives could encourage recommendations that benefit the bank rather than the customer.

Vélez says incentives and objectives must be designed around long-term customer satisfaction and believes his firm would seek to avoid conflicts. He argues Nubank need not own a foundation model to design its application’s behavior. These are management intentions and technical opinions, not proof that conflicts or errors can be eliminated.

He says starting a bank in Brazil without being Brazilian or having a banking and engineering background forced him to find people who filled his knowledge gaps. He values listening, questions and a beginner’s mind while building products and making decisions.

For a new company, he considers the early team, values and culture among the most consequential choices. He emphasizes customer focus, diversity of thought and a partnership culture. He credits those foundations with helping later decisions, an interpretation based on his own experience.

His product-strategy regret is entering investments through an acquisition rather than building organically. He says integration was harder than expected and the decision was influenced by market momentum. In hindsight he would have accepted a slower entry and now has a stronger preference for organic building.

Vélez defines success partly as undertaking a difficult adventure with strong colleagues, even if the business fails. He says building a bank from scratch was the hardest challenge he could imagine, which made the journey appealing to him.

In philanthropy, he weighs the desire to maximize long-term impact through education and leadership against urgent needs such as food and medical care. He says he has not resolved the tension and sees merit in both immediate help and systemic approaches.

On raising children with wealth, Vélez says he and his wife try to teach responsibility and provide age-appropriate challenges. He acknowledges they are still working out their approach. His view about struggle and confidence is a parenting perspective, not a universal rule.

In a quick-fire round, he says he would bring experienced people into Nubank sooner while retaining a questioning mindset. He recalls a 2017 proposal to change card-settlement timing that he feared would create a major funding need; he says the company and customers pressed regulators and the proposal did not proceed as feared.

He says the 2021 IPO benefited from preparation and timing before the following market correction. After the share price fell, the team tried to focus on execution rather than external interest rates and inflation. His account of timing is retrospective, not a method for predicting markets.

At the time, Nubank used a hybrid approach: teams spent at least one week in an office every six to eight weeks, with some choosing more. Vélez says this balances in-person coordination and access to global talent, and that the company was watching productivity and engagement.

For a hypothetical dinner, he chooses Steve Jobs, Karl Marx and Friedrich Nietzsche to ask about entrepreneurship and ideas. He also reveals that he enjoys singing opera despite considering himself a poor singer.

Looking toward 2028, Vélez hopes Nubank will lead in its three main Latin American markets, expand geographically and develop a broader consumer platform with AI assistance. These are the founder’s ambitions and predictions, not outcomes established in the interview. The guests close with thanks.