Primero Raises US$12 Million to Scale Enterprise AI in LATAM

Primero, a Latin American AI transformation startup, raised a US$12 million seed round led by Kaszek and General Catalyst to expand its Primia platform, which connects fragmented corporate systems and business rules with AI agents so large companies can automate processes while retaining control over their data and technology stack.

Primero is positioning the next stage of enterprise AI adoption in Latin America around a problem that goes beyond access to AI models: the difficulty of connecting those models to the complex systems, data, and business rules that govern large corporations. The Mexico-based startup announced a US$12 million seed round led by Kaszek and General Catalyst, with participation from Definition, 8VC, Conviction, and angel investors linked to companies including FEMSA, Bimbo, Kimberly-Clark, and Aeroméxico.

Primero says the funding will support talent growth, operations and expansion across the region. The company has already opened offices in Bogotá and Madrid and plans to continue expanding its presence in Latin America.

José Murillo, Co-CEO, Primero, says the company was created to address the gap between using individual AI tools and deploying AI across an organization. “There is still a very large gap between using an AI chat tool and having AI implemented in the company,” says Murillo.

Connecting AI to Fragmented Corporate Ecosystems

For large companies, the challenge is often not a lack of data but the difficulty of making that data usable across different systems. Primero works with corporations that can have more than 20,000 employees and hundreds of separate systems, including enterprise resource planning (ERP), customer relationship management (CRM), warehouse management systems (WMS), financial platforms, and spreadsheets.

According to Primero, the problem is particularly relevant in Latin America, where many large companies continue to operate legacy technology that was not designed to interact with modern AI systems. The company cites ManageEngine data showing that 87% of companies in Mexico say technology fragmentation limits their ability to scale AI initiatives. Globally, organizations manage an average of 957 applications, while only 27% are connected, according to MuleSoft’s 2026 Connectivity Benchmark Report.

Primero’s answer is Primia, a platform designed to operate as an intelligence layer above existing corporate systems rather than requiring companies to replace them. The platform connects to existing systems to read and write information while leaving the underlying databases in place. It also maps business information and rules that may not be formally documented but are necessary for employees to interpret and act on corporate data.

Murillo compares the problem to navigating a large city without a map. An ERP can provide the infrastructure and information, but AI still needs a way to understand where relevant information resides and how the organization uses it. One example is a company database containing dozens of columns that could potentially represent sales. Primia establishes canonical definitions for business objects, such as a customer or a sale, and determines which sources should be treated as authoritative.

That creates what Primero describes as an organizational map that AI systems can use to interpret corporate data.

From Data Integration to AI Agents

Primia is built around an ontology that combines information from multiple systems with business rules and organizational context. Those rules can include processes that are not documented in software but remain embedded in employees’ knowledge. Primero seeks to translate that institutional knowledge into a structure that AI systems can use when performing tasks.

On top of that layer, the company develops business intelligence modules, reports, and AI agents capable of executing specific processes. The approach is designed to allow companies to use different AI models without tying their corporate infrastructure to a single provider. Murillo says Primia can operate with models from companies such as OpenAI and Anthropic, among others. That model-agnostic approach is intended to give companies greater flexibility as AI models change in performance, cost and capabilities.

Primero also differentiates its model from traditional technology consulting. Rather than selling individual implementation projects based on hours worked, the company operates as a software platform with a licensing model. Murillo says this distinction is important because AI technology is evolving rapidly. A platform-based model, he argues, creates an incentive to continuously update the product as new models and capabilities become available.

The company also embeds engineers with customer teams. According to Primero, this allows its engineers to understand the technological environment, processes and organizational context required to implement AI in large corporations.

Early Deployments Target Measurable Business Outcomes

Primero says it has been operating for approximately one year and three months and is already working with companies including Kimberly-Clark de México, Verde Valle, SmartFit, Sede Café, Aurora Seguros, and the Mexico City Ministry of Economic Development. The company also works with organizations in financial services, consumer packaged goods, manufacturing, retail, and other sectors.

One of its use cases involves accounts receivable and retail deductions for consumer goods companies. These organizations can have employees manually review invoices, payments, warehouse deliveries and promotional agreements to determine whether deductions made by retailers are valid.

Primero says its AI systems can automate the collection and analysis of those documents and support the dispute process. In one customer case, the company says Primia identified more than US$10 million in deductions that should not have been applied, allowing the organization to pursue recovery of that money. The company says another deployment in the construction sector automates 85% of a process that previously required manual review, leaving 15% for human validation.

The objective, according to Primero, is to begin with use cases that can demonstrate financial returns and then expand toward broader AI transformation within the organization. This approach also reflects how the startup enters large companies. Murillo says there is often an internal executive or employee already looking for ways to apply AI, but that person needs a business case capable of demonstrating value to senior leadership.

Primero works with that internal champion to identify an initial process where AI can produce measurable returns and then uses the results as a foundation for broader adoption.

Latin America Becomes Part of the AI Infrastructure Race

Primero is focusing exclusively on large corporations, particularly organizations with complex technology environments. Murillo says many of these companies are family-owned businesses with a newer generation of leadership interested in technology and AI.

The company sees the region’s legacy infrastructure as both a challenge and an opportunity. According to Murillo, many Latin American corporations did not undertake the same level of cloud and software modernization seen in other markets over the past decade. Some still operate systems dating back decades, increasing the technical complexity of connecting them to modern AI.

Primero has therefore prioritized a local engineering presence. The company says it is expanding its team and has grown its workforce fivefold since the beginning of the year.

“While companies in the region adopt AI, Primero is building the fundamental infrastructure that will make that transformation possible,” says Alex Tran, Managing Director, General Catalyst.

For Primero, the next stage is to scale that infrastructure across more corporations and geographies. The company says the new capital will be directed primarily toward talent and expansion as it continues building its presence across Latin America.