# The Mini Guide to Vertical AI

#### Learn to spot opportunities
#### Greg Isenberg
Oct 10, 2024

Let’s talk about a space that could outgrow SaaS: vertical AI.

Vertical AI is software built for a specific industry’s quirks. Think healthcare, insurance, construction—each has its own workflows, rules, and pain points.

These quirks aren’t problems—they’re opportunities. Solve them, and businesses will pay big.

Imagine an AI that doesn’t just help an insurance agent but becomes the agent—underwriting policies, handling claims, spotting fraud. If you ran that business, you’d pay a fortune for it.

_via Greylock_

The rise of vertical software has been wild. [A16z](https://a16z.com/) breaks it down into 3 waves and it really clicked for me when they said it like this:

First wave, cloud brought services online (commerce with Shopify, service backoffice with ServiceTitan etc).

Second wave added fintech to cloud. This supercharged revenue (Toast hit $1.5B ARR, 80% from financial services).

Now, we're entering the third wave, cloud + fintech + AI. This is turning labor into software.

_via a16z_

So, I put together a quick guide to help you spot vertical AI startup opportunities.

##### Data goldmines
Look for industries drowning in unstructured data. Think doctor's notes, legal documents, or sensor data from factories. The messier and more proprietary, the better.

##### High-stakes decisions
Find areas where mistakes cost millions. Insurance underwriting, drug discovery, or supply chain management. AI that reduces errors here is invaluable.

##### Time-sucking tasks
Identify workflows where highly paid professionals spend hours on repetitive work. Legal discovery, financial audits, or medical diagnoses are prime targets.

##### Regulatory mazes
Sectors with complex, ever-changing regulations are ripe for AI. Think finance, healthcare, or environmental compliance.

##### Industry jargon
Fields with their own jargon and context are perfect. The more specialized, the harder it is for general AI to compete.

##### Fragmented industries
Sectors full of small players—construction, agriculture, local government—are gold mines lacking cutting-edge tech.

##### Data network effects
Look for places where more usage is better data, creating a flywheel.

##### Human-AI teamwork
Areas where AI can augment rather than replace humans entirely. I noticed this often leads to faster adoption which is obviously something we'd love.

If you tick a lot of these boxes, pretty sure you might be onto something big.

_loved this graphic which shows a bunch of vertical AI ideas by_

_via Greylock_

I hope this got your creative juices flowing.

What’s cool is the next wave of tech success stories won't all come from Stanford dorms. They'll emerge from law firms, hospital wards, and factory floors. Trapped in those people's brains is some insights that could lead to the next big vertical AI agent.

The vertical AI boom is just starting. The opportunity is massive, and the least competitive, most niche vertical AI software ones are the ones to look out for.

Time to start digging.