Wolox Co-Founder Returns to Silicon Valley with a Four-Person AI Startup — and Says Everyone Is Starting from Zero

How Luciana Reznik Went from a 400-Person Company to a Team of Four

Luciana Reznik spent a decade in Silicon Valley building Wolox, a digital transformation firm she co-founded with Nicolás Magni and Guido Marucci Blas. When Accenture acquired Wolox in 2021, Reznik had already led an organization of 400 people. Now she is back at the starting line — this time with Cedalio, a company whose entire team consists of the three founders and one senior engineer.

Cedalio uses AI agents to automate financial tasks for large companies. The platform retrieves invoices, extracts information, runs verification checks and loads the data into a client's management system so that administrative and treasury workflows can continue. The idea emerged from a narrower starting point: the company originally focused on data traceability, but while automating energy consumption data for enterprise clients, customers began asking whether the same system could handle all their invoices. About a year ago, Cedalio relaunched with that broader focus.

According to Reznik, the company works with more than ten large clients in the region, including Banco Galicia and Supervielle. She estimates the platform saves 80% of the time dedicated to invoice processing and that fewer than 5% of invoices require human intervention — figures reported by the company, whose real-world impact depends on each client's implemented processes. Cedalio raised an early funding round and went through Y Combinator, with Guillermo Rauch, founder and CEO of Vercel, among its investors. Reznik says the company has not sought additional financing since that initial round.

Reznik describes the current moment in Silicon Valley as one where even seasoned entrepreneurs are relearning fundamentals. "Today we are all at zero," she says, pointing to a dynamic where established founders and newcomers in their early twenties exchange knowledge in both directions. In her view, AI has lowered some of the barriers that historically constrained founders — particularly the level of capital and investment required to build products with meaningful reach.

What Cedalio's Bet on AI Agents Reveals About the New Startup Playbook

Why a Four-Person Team Is Now a Credible Structure

Reznik's comparison between Wolox and Cedalio is not just a personal anecdote — it illustrates a structural shift in how software companies can be built. At Wolox, she managed 400 people; at Cedalio, she spends more time on product and customers and less on administering a complex organization. The enabling factor is AI: agents handle the repetitive extraction, verification and data-entry tasks that once required dedicated operational staff. For context, this mirrors a broader pattern across enterprise software, where AI tooling is compressing the headcount needed for tasks like document processing, customer support and data reconciliation. The result is that a small, well-prepared team can now deliver products whose scope previously demanded a much larger structure.

Where Cedalio's Revenue Logic Actually Sits

Cedalio's business model targets a specific, high-friction task inside large organizations: opening supplier documents, reading amounts, cross-checking data and moving it between systems. The company says it works with more than ten large clients, including Banco Galicia and Supervielle. The value proposition is straightforward — reduce manual workload and flag exceptions rather than reviewing every invoice end to end. Reznik cites examples such as discrepancies in VAT calculations or problems validating an invoice with ARCA, Argentina's tax authority. The platform also enables downstream queries: expense reports by supplier, payment forecasts for the coming days. The commercial opportunity lies in the fact that this is work that consumes hours of skilled administrative time, and the clients themselves are asking for more automation, not less.

The Employment Question Cedalio Cannot Fully Answer

Reznik says no client has laid off staff because of Cedalio's platform, and that teams shift toward analysis, contract review and supplier conversations. But she adds a nuance: some tasks previously outsourced to third parties can now be handled internally. That observation describes what happens among Cedalio's own clients — it does not predict the effect of AI on administrative employment broadly. Still, it points to a real dynamic: automation may not eliminate headcount directly, but it can reshape where work is done and who captures the value. For context, this pattern — internalizing tasks that were once outsourced — has historically appeared in other automation waves, from accounting software to cloud infrastructure, and it tends to compress demand for external service providers before it affects internal teams.

Capital Still Matters — Just Differently

Reznik argues that starting a company today requires less capital than before, but she does not dismiss external funding. In Cedalio's case, selling to large organizations can take a year or more to close a first client. A lean structure reduces the cost of surviving that period but does not eliminate it. She also distinguishes between two goals: building a good business with limited resources and pursuing much greater scale within a short competitive window. For the latter, she believes being capitalized remains an advantage. Cedalio raised an early round and went through Y Combinator, with Guillermo Rauch among its investors, and has not sought additional financing since — a position that is only viable because the team is small and the product is already generating revenue from enterprise clients.

What Comes Next for Cedalio

The company's next steps combine geographic expansion and new products. Reznik describes a consolidated operation in Argentina, early clients in Chile and a government project in Central America, with plans to expand in Brazil and other regional markets. In parallel, Cedalio is working to add payment execution through partnerships with local financial institutions — a feature clients have requested after receiving information about their obligations. The vision is that agents will eventually operate with budgets and execute payments within defined rules and limits, though that requires resolving identification, authorization and regulatory compliance questions in each country. Reznik is explicit that not everything should be agentic: some tasks remain more efficiently handled by traditional software. The challenge is combining both approaches and deciding which responsibility belongs to each.

What Founders and CFOs Should Take from Cedalio's Approach

AI has lowered the capital barrier to building enterprise software, but it has not eliminated the time cost of selling to large organizations — and that distinction should shape how founders plan their runway.

• For founders building AI-native products: Reznik's experience shows that a team of four can now deliver what once required hundreds. But closing a first enterprise client can still take a year or more. Budget for that sales cycle even if your product is lean — Cedalio's early funding round and Y Combinator participation provided the runway to survive it.
• For CFOs and finance teams at large companies: Cedalio's model — automating invoice retrieval, verification and data entry, with human intervention only for exceptions — is a concrete example of where AI agents are already operational in enterprise finance. If your team spends hours on manual document processing, the build-versus-buy question is now more relevant than it was two years ago.
• For investors evaluating AI startups: Reznik's distinction between building a good business with limited resources and pursuing rapid scale is a useful filter. The first requires discipline; the second requires capital and a short competitive window. Ask which game the founder is actually playing.
• For regional fintech and enterprise software players: Cedalio's expansion into Chile, Central America and Brazil, plus its plan to add payment execution through local financial institution partnerships, signals where AI-driven financial automation is heading in Latin America. The regulatory complexity of payment execution — identification, authorization, compliance — is the bottleneck to watch.

Metrics to watch: Cedalio's client count beyond the current ten-plus, its progress on payment execution partnerships, and whether the company raises another funding round as it expands into Brazil.