AI Software

Sector thesis

AI Software is the layer of programs and tools that let companies actually *use* AI to solve real problems—think chatbots, forecasting tools, document analysis, and customer service automation. It's distinct from the chips and data centers (which we cover separately) that power AI; this is about the software sitting on top. Right now, this sector is interesting because companies are moving past the "let's experiment with AI" phase and into "how do we make this part of our actual business?" That shift is the real megatrend. Every industry—healthcare, finance, retail, manufacturing—is asking: where can AI cut costs or unlock new revenue? That creates steady demand for software companies that can deliver answers. Within AI Software, there are three main flavors. First: *enterprise AI platforms*—software that large companies license to build and run their own AI applications (think tools for banks or manufacturers). Second: *vertical AI*—pre-built solutions for specific industries, like AI for legal document review or medical imaging. Third: *consumer/productivity AI*—tools individuals and small teams use daily, often subscription-based. The biggest risk is that AI software is crowded and competitive. Margins can compress fast if a bigger player enters your niche. There's also execution risk: many AI companies are still proving they can turn hype into actual profit. And regulatory uncertainty—governments are still figuring out AI rules, which could affect pricing or who can sell what. For a retail portfolio, this sector works as a growth holding if you have a 3–5 year horizon. Watch for companies showing *actual customer adoption* (not just trial users) and *path to profitability*. Look at whether they're signing long-term contracts, not one-off pilots. This isn't a "set and forget" bet; you're betting on execution.