Marcos
Martinez

The Argentine engineer who quietly became Wall Street's AI whisperer

Before Marcos Martinez's company made the WealthTech100, before Broadridge Financial Solutions came knocking, before Mastercard and Wells Fargo showed up on the client list - there was a gift recommendation algorithm built on physics equations. Fligoo began life in 2011 trying to solve the problem of what to buy people for Christmas. What it actually built was a proprietary framework for inferring what humans want from the signals they leave behind in data.

Martinez, who was seven years into a software engineering degree at Universidad Empresarial Siglo 21 in Córdoba when he co-founded the company in October 2012, understood something that would take the financial industry a decade to accept: behavioral data is the most underused asset on the balance sheet. Banks sit on oceans of transaction history, engagement patterns, and demographic signals. Most do nothing with them. Fligoo built the machinery to turn that ocean into personalized, predictive intelligence at scale.

The pivot from social-graph gifting to financial AI was not a pivot at all in the way Martinez sees it. The core technology - algorithms drawn from statistics and physics that model consumer behavior - was always the point. The application changed. The ambition did not.

From gift engines to financial intelligence: the long game

Córdoba, Argentina's second city and academic hub, produces engineers the way Seattle produces software. Martinez studied there, built there, and still has a team there. It is where Fligoo's technical roots run deepest, even as the company planted its commercial flag at 44 Tehama Street in San Francisco, the kind of address that tells investors this is not a hobby.

The original FreshFeed product - visible even in the early Twitter handle @freshfeedapp - was a social news discovery tool, another early experiment in algorithmically surfacing what people want from the noise of the internet. These were not detours. They were the laboratory. By the time Fligoo turned its model toward financial services, Martinez and his co-founders had already run thousands of iterations on the fundamental question their platform answers: given everything we know about this person, what should we say to them, and when?

For wealth management, that question is extraordinarily valuable. A financial advisor managing hundreds of clients cannot realistically know which ones are thinking about retirement this week, which ones just changed jobs and need to revisit their portfolio, and which ones are quietly unhappy enough to take their assets elsewhere. Fligoo's AI can. Its platform ingests the signals - login patterns, document engagement, inquiry history, market events, life-stage indicators - and surfaces actionable intelligence to the advisor before the client picks up the phone.

This is not a product for startups. It is infrastructure for institutions. Martinez built it accordingly, on a stack that includes Apache Kafka, Amazon Kinesis, Amazon S3, AWS Glue, Redshift, OpenSearch, and Spark - a data pipeline architecture designed for institutional data volumes, regulatory environments, and zero tolerance for downtime.

The partnership that changed the conversation

In January 2021, Broadridge Financial Solutions - one of the most important infrastructure companies in global finance, processing over $10 trillion in securities transactions annually - announced a strategic partnership with Fligoo. The announcement ran on PR Newswire. It was the kind of institutional validation that does not come from a pitch deck. It comes from a working product that enterprise risk committees have approved.

Broadridge distributes to wealth management firms across North America. The Fligoo integration meant those firms could bolt predictive AI onto their existing workflows without rebuilding their technology stack. For Martinez, it was proof that the infrastructure bet had paid off. He did not build a prototype. He built something that Broadridge could trust with its clients' clients.

The client list that followed reads like a financial services directory: Mastercard, Wells Fargo, Cetera, Bancorp. Each represents a different segment of the market - payments, retail banking, independent wealth management, community banking. Fligoo's platform had to be flexible enough to serve all of them from a single codebase. That is a hard engineering problem, and it is the one Martinez spent the better part of a decade solving.

The CTO in the room

Martinez's engineering background is not incidental to Fligoo's success - it is structural. The company's technical decisions reflect someone who has written production code, not just approved architecture diagrams. The platform uses physics-based statistical modeling at its core, an approach that traces back to the original algorithm the team built for gift recommendations. That lineage matters. The models are not black boxes layered over existing tools. They are proprietary frameworks built from first principles, which is why they perform differently from commodity AI vendors in financial services contexts.

He holds a B.S. in Software Engineering from Universidad Empresarial Siglo 21 - a degree he completed in 2014, two years after formally joining Fligoo as CTO. He also holds a Diploma in PMI Project Management from PCI College, completed in 2011, the same year Fligoo was founded. In retrospect, the overlap looks intentional: he was running a company and finishing his education simultaneously, doing what builders do when the opportunity arrives before the credential.

Beyond Fligoo, Martinez has mentored entrepreneurs through the Entrepreneurship and Innovation Center at his alma mater. He is also a member of the Global CTO Forum, the peer community for technology executives navigating the scale-up phase. Both reflect a consistent pattern: he shares what he has learned rather than guarding it.

The long arc

The physics-to-finance thread

What makes Martinez an unusual figure in AI for finance is where the algorithm came from. Most fintech AI is built by teams that started in finance and bolted on machine learning. Fligoo came from the other direction: a team of engineers and data scientists who started with the science of human preference - borrowing from physics and statistical modeling - and then found that financial services was the highest-value application of that science.

That origin story shows up in the product. Fligoo does not describe itself as a CRM enhancement or a marketing automation tool, even though it functions as both. It describes itself as an intelligence layer - something that sits beneath the existing systems and makes them smarter. The difference matters to the CTOs and compliance officers who have to approve enterprise technology purchases. A smarter CRM is a tool. An intelligence layer is infrastructure. Infrastructure gets longer contracts and deeper integrations.

There is also something to the fact that Fligoo maintains a substantial engineering team in Córdoba. Argentina has one of Latin America's strongest software engineering talent pools, and Córdoba in particular has a university ecosystem that produces strong technical graduates. Martinez has the advantage of deep local knowledge - he studied there, built his network there, and can recruit there. The result is a company with San Francisco commercial ambition and Córdoba engineering depth, with additional presence in New York and São Paulo to cover the Americas' financial centers.

Fligoo's accelerator and investor network reflects this geographic range: MassChallenge gave it an early startup platform in the US, Plug and Play Japan extended it into Asian financial markets, FinTech Hive Accelerator connected it to Middle Eastern banking, and Plug and Play São Paulo anchored the Latin American institutional opportunity.

The details that don't make the press release

Martinez was finishing his software engineering degree while running a funded company. He completed his B.S. in 2014 - two years after formalizing his role as CTO of Fligoo. Not unusual for a certain kind of builder: the kind who starts before they're supposed to.

The very first Twitter account associated with Fligoo was @freshfeedapp - named for an early product called FreshFeed, a social news discovery tool that predates the financial pivot. The account survived long enough to appear in business intelligence databases as the company's Twitter handle. It is a small archaeological artifact of a company that has iterated its way to a very different place.

Fligoo's original algorithm for gift recommendations was grounded in physics and statistical mechanics - the same mathematical language used to model particle behavior. The intuition: human preference aggregates in ways that rhyme with physical systems. Whether that framing is literal or metaphorical, it produced a recommendation engine that worked well enough to attract institutional interest in the underlying method.

The company spans three continents - North America, South America, and through accelerator partnerships, Asia and the Middle East - from an 120-person team built by a co-founder who started in Argentina's interior and never stopped thinking in global terms.