Verticalized AI

Thoughts on why verticalized AI will win when it comes to the enterprise

Welcome to Infinite Curiosity, a weekly newsletter that explores the intersection of Artificial Intelligence and Startups. Tech enthusiasts across 200 countries have been reading what I write. Subscribe to this newsletter for free to receive it in your inbox every week:

---

Hello friends,

AI-infused applications are being built at a rapid pace. They are changing the way companies operate. They are helping companies to save time, money, and resources. One of the most promising avenues of AI is verticalized applications.

These are AI-infused products that are designed to solve specific problems in a particular vertical. It could be finance, healthcare, insurance, retail, education, material science, and more. LLMs are trained on all the data available and can do a large number of tasks. Verticalized AI models are specific to that domain and cannot really do much outside of that domain.

So why bother with verticalized AI? Don’t we want a single AI model that’s omniscient? Not when it comes to enterprise.

When it comes to enterprise software, here are 8 reasons why verticalized AI is poised to win big time over general-purpose applications:

1. Luxury of dealing only with domain-specific data

Verticalized AI products are able to take advantage of the specific data available in that vertical. This allows them to learn and improve more quickly than general-purpose AI applications. They have the luxury of only worrying about the domain-specific data. This is because they are designed to solve problems in a particular vertical, where there is a wealth of data available. Do we really care if a healthcare AI model doesn’t know how to detect credit card fraud? Nope.

2. Infusing domain knowledge into the product

Verticalized AI products can be tailored to the specific needs of that vertical by infusing the product with domain knowledge. This means that they can be more effective at solving problems and improving outcomes. This knowledge can be used to train AI models and to improve their performance. If I’m a lawyer using AI for my work, I need it to know the basic terminology of my domain. And I need it to know exactly how I work.

3. Domain expertise of the people

Verticalized AI products are usually developed by people with deep knowledge of how their industry works. They are designed to solve problems that are not super obvious to someone outside the domain. This domain expertise is essential for developing AI applications that are effective and useful.

4. Integrating with existing systems

Verticalized AI products can be more easily integrated with existing systems and processes within a domain. This makes it easier for businesses to adopt and use them. Each industry has its own set of requirements. And they have a set of systems and processes. Products that can integrate well with that have a much higher likelihood of adoption as compares to generic applications.

5. Customer willingness to pay is high

Large companies are happy to pay 100x more for domain-specific solutions vs generic tools. Why? Because they don't want to figure out how to stitch together many different tools to build a full solution. Even if those tools are the best in the world at what they do.

6. Value capture is high

You can become part of the customer's workflow. The ability to capture the value you're creating is higher for verticalized AI products. You can charge a premium for building a complete solution for the enterprise. Why? Because it will cost way more for the customer to go out and use disjointed tools.

7. Intra-domain network effects

If you become the go-to product within a vertical, everyone in the vertical will flock to you and stay with you. It's hard to unseat the #1 vendor in a specific vertical. The mindshare of the #1 product within any vertical is high.

8. Snowball effect of product usage data

As more customers interact with your product, that usage data becomes your moat. That data can be fed back into the AI model to make it even better. This snowball effect will continue to grow. This is achieved using reinforcement learning.

Verticalized AI is particularly well suited to the enterprise. And companies that build verticalized AI applications are poised to win big time.

If you’re a founder building a verticalized AI product, I’d love to hear from you. You can connect with me on LinkedIn or Twitter. DMs are open.

---

If you are getting value from this newsletter, consider subscribing for free and sharing it with 1 friend who’s curious about AI: