MV aims for R$1 billion by 2027 by turning hospital SaaS into an AI platform

The trajectory of Paulo Magnus, founder and CEO of MV, began with a mundane operational detail: employees at a hospital in Rio Grande do Sul spent hours trying to decipher doctors' handwriting to turn patient care into billing. Four decades later, that same friction, between what happens on the front line of healthcare and what becomes revenue in the back office, is the engine of a company that generated more than R$640 million in revenue in 2026 and projects crossing R$1 billion in 2027, according to a report by Exame.

For those thinking about B2B scale-up strategy, the case is interesting less for the round number and more for the path to it: four decades without knocking on VCs' back door, slow consolidation of a vertical market, and now, a shift from a single product to a portfolio of solutions with embedded AI.

From outsourcing billing to selling software

MV's business model was born as a service, not a product. Magnus proposed to a hospital in Torres (RS) that he take over billing in exchange for 2% of the amount received, plus equipment provided by the institution itself. That's how he financed the initial operation, in a logic we would today recognize as an MVP paid for by the pilot customer itself.

The break toward product came in 1992, when the company launched an integrated hospital management system combining administrative and care processes. Instead of selling hours of work, MV started selling the standardization of that work as reusable software, the classic transition from service provider to scalable technology vendor.

In short: the manual service validated the problem and generated cash; the product scaled the solution. That's the sequence missing in many startups that try to jump straight to a platform without first understanding the process they're automating.

The numbers behind the R$1 billion target

MV does not disclose external investment rounds in the report, which suggests mostly organic growth over nearly 40 years, a trait rare among technology companies now aiming for a billion in revenue.

The 2026 investment goes into three fronts: modernizing legacy systems, automating the revenue cycle, and AI tools for summarizing clinical information. Today MV's customer base totals more than 5,600 healthcare institutions in 13 countries, covering about 178,000 beds, according to the company's own data cited by Exame.

SOUL Agents, Maida Health, and Luiza Health: a platform instead of a single product

Instead of stacking features into a single piece of software, MV is slicing the problem into specialized brands. SOUL Agents automates workflows and supports decisions within the hospital operations that customers already run on MV's systems, a bet on revenue expansion within the installed base, cheaper than winning a new customer.

Maida Health, another company in the group, serves health insurers and self-managed health plans with data analysis and care regulation. According to the company, the technology automated about 20,000 medical authorization reviews in three months in a single deployment, a figure that any board evaluating AI ROI in healthcare back-office operations should ask to benchmark.

Luiza Health, still in preparation, will cross-reference clinical data with equipment information to help teams identify signs of deterioration in hospitalized patients. Magnus is explicit about the limit of automation here: decisions about clinical conduct remain with healthcare professionals, not the algorithm, which reduces the product's regulatory and liability exposure.

For healthtech leaders, this division into separate brands (SOUL Agents, Maida Health, Luiza Health) is as much a governance decision as a product one: each unit can raise capital, close a partnership, or even be sold separately, without compromising MV's core.

What this teaches B2B scale-up leaders

Three decisions from Magnus's trajectory serve as evaluation criteria for other vertical SaaS founders aiming to become a platform:

• Validating the manual process before selling the software that replaces it, instead of assuming the customer's pain point;
• Using a concentrated talent hub (in this case, the Universidade Federal de Pernambuco, a public university in Recife) as a hiring competitive advantage, not just lower payroll cost;
• Slicing growth into distinct products/brands as the customer base matures, instead of inflating a single monolith that becomes expensive to maintain and hard to sell in pieces.

MV's entry into São Paulo at the turn of the 2000s, driven by the system overhauls needed for the Y2K bug, is also a reminder that external regulatory or technological windows (compliance, mandatory migration, new law) tend to be a more effective sales trigger than any marketing campaign.

Where the model could stall

The most fragile point of Magnus's thesis isn't technological, it's institutional. He argues that patient data should follow the patient across different healthcare services, which requires interoperability between competing systems, not just within MV's own base.

I would very much like institutions and healthcare services to decide in favor of interoperability. We work to get the data to the person. That data turns into a clinical summary when they go to the doctor.

Paulo Magnus, founder and CEO of MV

The problem is that this decision isn't in MV's hands: it depends on hospitals, insurers, and regulators in 13 countries agreeing to exchange data between different platforms, many of them direct competitors. Without that, the promise of AI that summarizes a patient's clinical history stays limited to what happens within MV's own ecosystem, which reduces the data's value precisely in the scenario where the patient moves between different institutions.

For anyone evaluating an investment or partnership with a healthtech at this stage, the right question for the CTO isn't "does the AI work," it's: what percentage of the customer base has actually enabled interoperability, and how much of the R$1 billion target depends on regulatory adoption the company doesn't control.