CloudWalk's AI Army Hits $1.3B Run Rate with Just 720 Employees

SUNNYVALE, CA – March 11, 2026 – While most of Silicon Valley debates the future of artificial intelligence, one company has quietly built an AI-powered empire that is rewriting the rules of productivity and profitability. CloudWalk, the fintech firm behind Brazil's InfinitePay and the newly launched JIM.com in the U.S., closed 2025 not just with impressive growth, but with efficiency metrics that dwarf many of the world's largest tech titans.

The company announced it ended the year with $990 million in revenue, a 104% year-over-year surge, while crossing a staggering $1.3 billion annualized revenue run rate in December. Perhaps more remarkably, it achieved this with a lean team of just 720 people, translating to an industry-shattering $1.8 million in revenue per employee. Net income also soared 90% to $110 million.

These figures aren't the result of aggressive hiring or traditional expansion. They are the product of a radical operating model where autonomous AI agents are not just tools, but the primary workforce. As CloudWalk founder and CEO Luis Silva puts it, “Every fintech says they use AI. We are an AI company that happens to do finance.”

The Autonomous AI Engine

CloudWalk's staggering efficiency stems from its core philosophy: AI agents don't assist employees; they are the employees. While competitors use AI to suggest email responses or analyze data for human review, CloudWalk deploys a new generation of autonomous agents, built on proprietary large language models, to independently execute core business functions. Humans, according to Silva, now primarily “set policy, handle exceptions, and govern risk.”

This AI-native approach has yielded remarkable results. The company's revenue per employee of $1.8 million places it in the elite echelons of global tech efficiency, comparable to or exceeding giants like Alphabet ($1.91 million) and Meta ($2.19 million), and far surpassing many others. This metric is not a one-off spike; it’s the result of a deliberate, long-term strategy. The figure is more than double the company's own $796,000 RPE from mid-2024, demonstrating the compounding effect of its AI model.

Concrete examples of this model in action include:

• Claudio Walker: A customer service agent that autonomously resolves over 90% of customer issues for Brazil's InfinitePay. It analyzes historical and real-time data to provide responses in seconds, reportedly performing the work equivalent of 1,100 human staff members.
• Bela: A WhatsApp-based sales agent that handles 50,000 conversations a month, tripling conversion rates with its 24/7 availability.

Beyond customer-facing roles, these agents are at work behind the scenes, underwriting credit, building new software, creating marketing campaigns, and running a sophisticated fraud-prevention system that stopped an estimated $2 billion in fraudulent transactions between mid-2023 and mid-2024.

A Tale of Two Markets

CloudWalk’s strategy has proven uniquely effective across vastly different economic landscapes, fueling rapid growth in both the emerging Brazilian market and the mature U.S. market.

In Brazil, its InfinitePay platform has become a dominant force for small and medium-sized businesses. The active seller base more than doubled in 2025, growing from 3 million to over 6.3 million merchants across every municipality in the country. This explosive growth propelled InfinitePay to become the most downloaded finance app in the Brazilian App Store in December, boasting a near-perfect 4.9-star rating from over half a million reviews. This success was bolstered by significant regulatory milestones, including securing a full credit institution license from Brazil's Central Bank and earning a stellar 'AA-(bra)' rating with a Positive Outlook from Fitch Ratings. These achievements grant the company full autonomy over its lending operations and signal strong institutional confidence in its AI-driven business model.

Simultaneously, CloudWalk made a bold entry into the United States in early 2025 with JIM.com. Targeting micro-merchants and gig workers, JIM.com is far more than a simple payment app. It’s pitched as the first autonomous financial agent that works for the seller. The service turns any smartphone into a payment terminal with instant settlement, but its true power lies in its AI core. As Silva explains, “There is no dashboard. No manual. No learning curve. The agent runs the business and the merchant runs their life.” This means the AI can proactively diagnose and resolve declined payments, identify cash flow risks, and even build a functional website for a seller using only their transaction data. The concept resonated immediately, with JIM.com acquiring users in all 50 states within months of launch, primarily through word-of-mouth.

Building 'Self-Driving Finance'

CloudWalk's ultimate ambition extends far beyond its current products. The company is charting a course toward what it calls “Self-Driving Finance”—a fully autonomous financial system where AI agents manage payments, credit, and strategic decisions with minimal human oversight. This vision represents a paradigm shift from the current financial system, promising unprecedented speed, accessibility, and efficiency.

The journey toward this future is already demonstrating tangible benefits beyond the balance sheet. By automating complex financial tasks, CloudWalk is lowering the barrier to entry for entrepreneurs and gig workers who lack the time or resources for traditional business administration. The success of InfinitePay in Brazil and the rapid adoption of JIM.com in the U.S. highlight a massive, underserved market eager for tools that simplify the complexities of commerce.

This AI-first model has proven its ability to scale with incredible speed without the corresponding bloat in headcount that typically plagues growing companies. As the company continues its global expansion, its lean, AI-driven structure provides a formidable competitive advantage. “Most companies slow down as they grow. We doubled revenue with just 720 people,” Silva stated. “The model is proven. Now it scales.”