Latin America and Europe to Boston: What Separates the Startups That Scale

What we talked about.

In this episode of the Sunny Ray Show, host Sunny Ray talks with Greg Bauer, managing partner of Scale-Up Labs, a founder-led investment and venture builder firm working across Boston, Latin America, and Europe. Bauer traces his path from building three startups, including one he took public, to launching Spinnaker Venture Partners in 2008, which later became part of Scale-Up Labs. He explains how the firm both originates companies in its Boston venture builder and runs a yearly accelerator cohort of 10 companies selected from roughly 600 applicants. The conversation digs into cultural differences in how founders and investors treat failure, why Latin American and European investors often prize cash flow while US investors chase valuation, and why that gap can stall international expansion. Bauer also describes the firm's focus on AI native companies in health tech, advanced manufacturing, and energy and sustainability, his advice on raising capital in a more cautious market, and his five-year vision for the firm, including regional funds in Latin America and Europe.

Greg Bauer explains why cash flow thinking abroad and valuation thinking in the US decide which startups scale, and what AI native really means.

The questions, and the answers.

What has building next to founders taught you that no investor deck ever could?

I've learned how differently cultures treat risk and failure. In Latin America and Europe, failure carries a stigma with family, friends, and investors, which makes founders reluctant to take big risks. In the US, failure signals a willingness to try something significant. Some West Coast investors won't back a founder who hasn't failed at least once.

How do you find and back companies, and how do you work with them?

We invest two ways. We originate companies in our Boston venture builder, and we find companies in Latin America and Europe and put them through an acceleration program that prepares them to internationalize and raise capital in the US. We take a position when they enter, so we're almost like co-founders, very hands-on.

How many pitches do you see, and what gets through?

Our accelerator takes one cohort of 10 companies a year, and we review about 600 applicants to get there. For the investment fund, I get around 50 unsolicited emails a week, and maybe one gets really read. Many are in sectors we don't invest in, because people don't do their homework.

What do you mean by AI native, and which sectors do you focus on?

AI native means the company couldn't exist without artificial intelligence at its core, usually through specialized models or small language models. Wrappers around big language models aren't sustainable and carry little intellectual property. The company must solve a significant, investable problem. We only look at health tech, advanced manufacturing, and energy and sustainability.

What most often stalls a promising company between first customers and real growth?

Often founders aren't willing to make the sacrifices needed to scale internationally. They see the US as monolithic when it's countless markets and cultures. There's also an investor conflict: Latin American and European investors want cash flow and EBITDA, while US investors back valuation growth, and home investors sometimes block US-style capital.

If you could fix one thing about how early stage companies get funded, what would it be?

For international founders, I'd want them to raise toward a destination, not just the next funding event. For US companies, I'd want a deliberate capital formation strategy instead of spray and pray. Markets have also grown more cautious: nine months ago I'd say go raise $20 million, and today I'd say aim for $5 million.

Five years from now, what does Scale-Up Labs look like?

I hope our portfolio value is north of a billion dollars. We're raising our first regional fund for Latin America now, and I'd expect to be on fund three in Latin America and fund two in Europe by then. The pace of AI makes it hard to predict, but I expect far more possible than we can imagine today.