Many VC funds claim to be only investing in AI startups, or from the founder perspective, have the (false) belief that they can only raise capital if they call themselves AI companies. At BuenTrip Ventures, we think it’s time to interrogate these beliefs to better understand the sector and the technology, and to make better investment decisions in the face of the AI boom.
While AI has become a popular buzzword in startup and investment circles, the reality is more nuanced. AI is not a monolithic sector, but rather a technology that can be applied across a wide range of industries and use cases. In this article, we'll explore the various ways AI manifests - both as its own distinct sector and as a supporting tool within other verticals. Understanding these distinctions is crucial for making sound investment decisions in the current AI landscape.
There are moments when AI absolutely behaves like its own sector. These include:
Foundation models and core infrastructure: OpenAI, Anthropic (Claude), and Google DeepMind (Gemini) don’t just use AI—they are AI. Their entire business is built on creating and scaling general-purpose models that others can build on.
Hardware and compute stack: The massive demand for GPUs and AI chips makes companies like NVIDIA central players in this ecosystem. Cloud infrastructure and training pipelines are another foundational layer.
Training and data ops: From data labeling to self-supervised learning to synthetic data generation, the supporting scaffolding for AI models is its own high-growth industry.
Multi-agent systems and orchestration: Companies building multi-agent coordination layers, memory systems, and tooling for reasoning—like LangChain or emerging frameworks—are building the “middleware” of the AI stack.
User interphases and connectors: Think of chatbots, speech interfaces, or visual assistants. These are not just UX layers—they define how humans interact with machine intelligence.
This is what we might call "horizontal AI"—a set of startups that develop AI infrastructure, models, or tooling to other companies. In this view, AI resembles a sector, with its own supply chain, value chain, and economic logic.
But here’s where the confusion begins: a growing number of startups describe themselves as "AI companies" simply because they use AI.
This is the "vertical AI" fallacy—startups that are solving problems in real-world sectors (healthcare, education, logistics, FinTech) using AI as a tool, not as the product itself. Their success is far more dependent on their sector expertise, distribution channels, and regulatory navigation than on their use of AI.
BuenTrip Portfolio Examples
Mercately isn’t building a large language model. It’s not a chip manufacturer. What it is doing is integrating AI to replace outdated user interfaces for SMB eCommerce operators—helping them chat with customers, track sales, and increase conversion. AI enhances their product, but it doesn’t define the company or its success.
Other startups in our portfolio use AI to improve internal processes—code generation, document automation, customer segmentation—but none of these companies would cease to exist if the AI landscape changed tomorrow. They would adapt because their core value is not in developing horizontal AI, but in solving a pain point for a particular customer.
A Rare AI Startup
Now contrast that with Synthera, who help hedge funds and other financial institutions to manage risk synthetic financial market data. Synthera isn’t relying on third-party foundation models. They’re developing their own domain-specific models for financial analysis. Their first hires? Quant researchers, statisticians, and people with deep data engineering experience who can build models from scratch. Their hires are domain experts, not sales, marketing or product specialists. The team composition is a telling sign that the startup is structured in a different way than most startups that call themselves AI companies, but are hiring and building teams just like before the AI boom.
This is closer to horizontal AI—a rare case where the model is so domain-specific that the company becomes an AI startup, even if the market they serve is vertical.
Under this framework, it is unclear that Latin America needs thousands of AI-first companies. It needs companies solving real, rooted problems—with AI when it makes sense.
Take SCITRIX, a startup that uses a specialized AI model to analyze the chemical composition of pharmaceutical solutions. Previously, local pharma labs had to send samples to Europe for high-precision analysis—an expensive and time-consuming process. SCITRIX’s AI-driven solution dramatically reduces costs and turnaround times. But what sets it apart isn’t just the use of AI; it’s the combination of technical innovation with deep domain knowledge and an understanding of local market constraints. Unlike startups that simply layer general-purpose AI models onto existing workflows, SCITRIX integrates a specialized AI model as a core component of its product—developed with a clear, localized purpose in mind.
The next time you hear a fund declare, “We invest in AI,” pause for a moment. More often than not, that’s code for “we’re a generalist fund adapting to a new technological landscape.” Ask yourself: What does that really mean? There’s a big difference between:
Both approaches can yield returns—but they require very different expertise, portfolio construction logic, and risk tolerance.
At BuenTrip Ventures, we don’t chase buzzwords. We back founders solving real problems in real industries—healthcare, logistics, fintech, and education—with the right tools for the job. Sometimes, those tools include AI. But the best startups aren’t built on tools. They’re built on insight, execution, and community.
Saying “AI is not a sector” is as simplistic as saying “software is not a sector.” It depends on what you mean. There are clear cases when AI is evidently a sector, but others where it is behaving more as a buzzword.
What’s clear is that as the next wave of Latin American startups emerges, AI will be in the background of many of them. The winners won’t be those who chase trends, but those who understand their industries deeply—and know when, how, and why to apply new technologies.
