Brazil's challenge to the AI-native ERP thesis

Gonzalo Parejo is the founder and CEO of Kamino, a financial operating platform for mid-market companies in Brazil.

AI-native ERP startups are raising capital on the idea that AI can replace the spreadsheet as finance's default operating system. That may be true in the U.S., where spreadsheets are at the center of many finance workflows. But in Brazil and many other emerging markets, finance teams rely on ERP systems for a different reason: complying with complex tax and regulatory requirements. That changes both the role the software plays and the problems customers are willing to pay to solve.

Investors applying this thesis to Brazil are making a foundational error. The ERP was never the CFO’s operating system, it was designed to be a compliance bureau. That distinction changes everything about where AI gets applied and what category of platform actually wins. The answer is not a smarter ERP. It is a new category: the “CFO’s Operating System”.

Why the ERP was never the CFO’s Operating System

In Brazil, the ERP’s primary job is to calculate and transmit tax obligations across a fragmented federal, state, and municipal system, a tax regime estimated at up to ten times more complex than its American equivalent. The ERP does this with precision. 

But the picture it produces is designed for auditors and tax authorities, not for the CFO making cash decisions. A mid-market company commonly operates with multiple tax registrations across states and five to fifteen active bank accounts. The ERP books revenue on day one under accrual accounting, but actual cash may not settle for 30, 60, or 90 days. The ERP manages the “tax past.” The CFO needs to manage the “cash future.”

The telling detail: the primary tool these finance teams use to manage cash flow is not a purpose-built platform. It is Excel. As Andreessen Horowitz documented, finance teams stick with Excel because nothing integrates. The ERP handles compliance, the banks handle execution, and the managerial layer lives in spreadsheets.

The winning platform in these markets will not be an AI-native ERP. What’s missing isn’t a better layer on top of the ERP. What's missing is a single system that owns accounts payable, accounts receivable, reconciliation, and cash flow, where the ERP is one data source among several. This is the “CFO’s Operating System,” and it’s a category that simply doesn’t need to exist in the United States, where the ERP already does this job. But in any market where the ERP functions primarily as a compliance tool, much of Latin America, Southeast Asia, parts of Africa, no single system provides it. The CFO stitches together the operational truth by hand.

AI makes this gap wider, not narrower

Here is what many investors miss: An AI agent trained on ERP data in Brazil sees R$5 million (roughly $1 million) in booked revenue and recommends a hiring plan. What it cannot see is that most of that revenue sits in installment receivables, subject to tax withholdings that could reduce net cash by 15 to 20%. The cash won't arrive for 30, 60, or 90 days, but the AI has already made its recommendation. The result: a company that looks profitable on paper while walking into a liquidity crisis.

This is where AI flips from risk to advantage. Once the execution layer captures real-time, cash-basis data, AI can deliver on its promise: forecasting liquidity, flagging anomalies before they become crises, giving a three-person finance team the analytical power of a twenty-person department. The platform that captures the data owns the AI opportunity.

The opportunity

AI will reshape financial software; the global venture ecosystem is not wrong here. But the assumption that this will center on the ERP is a developed-market bias. In Brazil, the ERP is a tax machine, not a financial brain. The category-defining opportunity is the CFO’s Operating System: the financial operating layer that combines all relevant data sources into a single source of managerial truth. Whoever builds it owns the data, owns the AI advantage, and owns the moat. The complexity that makes these markets harder to serve is also what creates the opportunity. The winners will not be the companies rebuilding the ERP. They will be the ones that become the system CFOs actually run their businesses on.