Artificial intelligence is quickly becoming a core part of how new software companies are built. Instead of using AI as a simple feature, many startups are now designing their entire product around it. As the market grows, investors are looking for companies that can build strong technology, protect their data, and scale efficiently.
To understand how this shift is shaping early-stage investing, we spoke with Patricia Pastor, General Partner at Next Tier Ventures. Patricia focuses on backing founders who are building AI-native companies where the technology is at the center of the product.
In our conversation, she explains what makes AI startups stand out today, how investor expectations around AI funds have changed, and where she sees the most promising opportunities as the industry continues to grow.
Artificial intelligence is becoming the foundation of many new software companies. Instead of adding AI as a feature, founders are now building products where AI is at the core of the system. As this transition continues, investors are focusing on teams that can turn powerful technology into scalable and defensible businesses.
We asked Patricia Pastor about what first drew her to AI investing and what she looks for in founders building AI-native companies today.
What initially drew us to this space was the profound shift in technological power—the realization that AI isn’t just a tool, but a catalyst that redefines human potential at every level.
However, as the market matures, we’ve found that the real difference-makers are founders with a ‘Full-Stack Obsession.’ Being a model expert is a great start, but we back leaders who think deeply about infrastructure and operational excellence. We are particularly inspired by the vision of Sovereign AI: companies that empower enterprises to own their intelligence and their data.
We look for founders who help businesses move away from ‘renting’ their core technology and instead build it as a strategic IP asset. In an era where privacy is the ultimate gold standard, the winner is the founder who builds a fortress around the client’s data perimeter.
Is there an AI vertical or application you believe is meaningfully underappreciated by the market today?
Our thesis has evolved from exploring ‘AI as a feature’ to championing AI as the new foundational architecture of software. The world has moved past the stage of simple enhancements.
We are no longer looking for functionality that can be disrupted by a single model update. Instead, we invest in defensible architectures. We focus on companies where AI isn’t just an add-on, but the very DNA of the system—solutions that are robust, efficient, and built to thrive as the industry scales.
What Is The LP POV in This Sector?
Over the past few years, artificial intelligence has attracted a wave of investor attention. But as the market matures, limited partners are becoming more selective and are asking deeper questions about how AI companies create long-term value.
We asked Patricia how LP expectations around AI-focused funds have evolved and what new concerns or opportunities are shaping investor conversations today.
We’ve seen a healthy evolution from ‘AI curiosity’ to ‘AI maturity.’ Two years ago, the goal was simply to get a foot in the door; today, our LPs are looking for sustainable value.
They are asking the tough, necessary questions: ‘How do these companies stay resilient if the big labs slash their prices?’ or ‘How does the regulatory landscape, like the EU AI Act, become a competitive advantage rather than a hurdle?’
The most exciting part is that the capital is becoming more sophisticated. We are seeing ‘smart money’ that realizes the real prize isn’t just the model—it’s how the AI is woven into the workflow and fueled by proprietary data. It’s no longer about who has the AI, but who builds the most enduring value with it.
Early-stage Fundraising & Execution
For many AI startups, raising capital in the early stages poses unique challenges. Founders are often navigating rapid technological change while also building a sustainable business model.
We asked Patricia about the most common mistakes she sees from AI founders when raising capital and what truly differentiates strong teams at the pre-seed and seed stages.
The most frequent trap is losing sight of the business fundamentals in the rush for growth. In the early days, it’s easy to overlook unit economics, but we believe that traction is only meaningful if it’s sustainable.
We see many teams unintentionally letting their margins be squeezed by external compute costs without a clear roadmap for efficiency.
From an investor’s perspective, what really differentiates strong AI startups at pre-seed and seed today?
What truly sets a pre-seed or seed team apart today is Architectural Empathy. We are drawn to founders who are not just brilliant at coding, but brilliant at designing for scale and cost-control from day one.
The winners are those building proprietary assets—teams that know how to own their intelligence rather than just renting it. That strategic foresight is what turns a great project into a generational company.
The Role of Geography & Ecosystems
AI innovation is happening across multiple regions, each with its own strengths and challenges. While some ecosystems benefit from access to capital and talent, others are emerging as strong hubs for applied AI and specialized expertise.
We asked Patricia how she compares different regions for building AI companies and where she currently sees the strongest momentum.
You invest across Europe, North America, and LatAm. How do you compare these regions in terms of AI talent, ambition, and company-building maturity?
We invest in companies that are “born global,” but each region has its own DNA:
USA: Unbridled ambition and infinite capital, but often with absurd operational and talent costs.
Europe: Superior technical talent and a clear regulatory advantage.
LatAm Pure pragmatism. They know how to apply AI to solve real inefficiencies in traditional sectors with highly efficient budgets.
Are there specific markets or ecosystems where you’re currently seeing disproportionate momentum in AI?
The reality is that we are witnessing the first truly borderless technological revolution. Momentum isn’t just happening in specific cities; it’s happening wherever there is a high density of computer and specialized talent.
We’re seeing disproportionate momentum in vertical AI (health, legal, finance), where models are embedded directly into workflows and deliver measurable ROI. There’s also strong traction in AI agents for back-office automation and in AI-native cybersecurity.
Geographically, the most interesting ecosystems right now are SF/NY for agents and infra, Israel for cyber, and Southern Europe for applied B2B AI, where competition is lower and domain data is a real moat.
How do regulatory, cultural, or market structures shape how AI startups scale across these regions?
In Europe, we see that compliance with GDPR and the AI Act has shifted from being a hurdle to becoming a competitive advantage for Enterprise sales.
Global corporations (Banking, Insurance) prefer buying European software because it guarantees security and compliance “out of the box”—something that is sometimes missing in solutions from other markets.
Looking for Signal vs. Noise in AI
With artificial intelligence dominating headlines and attracting large amounts of capital, it can be difficult to separate real opportunities from hype. Investors are increasingly focused on identifying businesses that solve meaningful problems and can build long-term advantages.
We asked Patricia where she sees genuine opportunity in the AI landscape and which areas may be attracting more attention than they deserve.
The market is saturated with noise in the generic application layer (”wrappers”). For us, the ‘signal’ lies in two areas: Orchestration and Verticalization.
The genuine opportunity is in operational infrastructure. Poor base management and planning can kill a startup. That is why we bet on Intelligent Orchestration: tools (like Optiak in our portfolio) that act as a ‘GPS,’ deciding which model to use at any given moment to optimize costs and avoid vendor lock-in.
Is there an AI vertical or application you believe is meaningfully underappreciated by the market today?
One of the most underappreciated AI verticals is automating internal operations in traditional sectors (industry, logistics, non-clinical healthcare, public sector).
While the market focuses on generic copilots, the real value lies in replacing entire processes, leveraging proprietary data, delivering clear ROI, and creating high barriers to entry.
The next winners won’t be better chatbots—they’ll be systems that redefine how organizations operate from the inside.
Conversely, what parts of the AI landscape feel overcapitalized or premature?
Today, the market is clearly overcapitalizing horizontal copilots: productivity assistants, generic content generators, and “enhanced” chatbots.Many of these products are saturated, rely on easily replicable public data, and have low barriers to entry, making it hard to build defensible businesses.
Additionally, some highly futuristic AI applications, like AGI for complex tasks, are still premature: lots of hype, massive investment, and uncertain returns. In short, too much capital is chasing short-term promises instead of solving real operational problems inside companies.
Looking Ahead at 2026
As AI adoption accelerates, the next few years are likely to bring major changes in how companies build and deploy intelligent systems. New technologies, infrastructure, and business models are already beginning to influence the market.
We asked Patricia where she expects the strongest growth to happen in the near future and which risks founders and investors should pay closer attention to.
Which areas within AI do you expect to see the strongest growth over the next 12–24 months?
We are moving toward a world of Zero-Touch Intelligence. The next 24 months won’t just be about smarter apps, but about a fundamental shift in how the digital world breathes:
The Rise of Autonomous Infrastructure Agents: We are backing the birth of the ‘Self-Healing Enterprise.’ Imagine a world where AI doesn’t just assist humans, but independently orchestrates servers and neutralizes cyber threats in real-time, without human intervention. This is the end of reactive maintenance and the beginning of autonomous resilience.
The Edge & Local AI Consolidation: The future of AI isn’t just in the giant, distant cloud; it’s happening right where the action is. We are seeing a massive shift toward Edge Intelligence—small, hyper-powerful models running on private servers and local devices. By slashing latency and compute costs, we are moving from ‘Centralized Intelligence’ to ‘Sovereign Performance.’
The Bottom Line: At Next Tier, we invest in the architects of this transition. We are moving from AI that talks to us, to AI that works for us, silently and securely, at the very edge of the network.
As capital and competition accelerate, what risks, technical, commercial, or ethical, should founders and investors be paying more attention to?
We believe the AI ecosystem is entering a phase of much greater discipline. Two structural risks are becoming increasingly clear.
First, model dependency. The base model layer is evolving extremely fast and becoming increasingly commoditized. If a company’s only advantage is access to a specific model or API, it is very difficult to build a durable business. What we look for instead are companies that create defensibility above the model layer — through proprietary workflows, domain expertise, unique data, or deep integration into customer operations.
Second, compute economics. AI businesses are fundamentally constrained by infrastructure costs. Training and inference require significant compute, and for many startups the cost structure can quickly become unsustainable if it is not designed carefully from the beginning. As the market matures, we expect efficiency in architecture, model selection, and inference costs to become a major competitive advantage.
Beyond the technical layer, we also pay close attention to distribution and workflow ownership. In our view, the long-term winners will not necessarily be those with the most advanced models, but those that successfully embed AI into real business processes and become indispensable to their customers.