---
title: "Vertical SaaS AI agents · agents built for one industry · Agent Fieldbook"
url: https://agentfieldbook.org/catalog/vertical_saas/
description: "Vertical SaaS agents are built for one industry's workflow and sold as software: domain data, integrations and guardrails on top of a frontier model or platform. Depth over generality, each specced against primary sources and drawn as a topology specimen."
section: "Catalogue · Vertical SaaS"
source: Agent Fieldbook — generated from the published page
---

# Vertical SaaS agents

Agents built for one industry's workflow and sold as software: they know a domain's data, rules and jargon rather than trying to be general. Depth over breadth. Most run on a frontier model or platform underneath and win on proprietary data, workflow fit and compliance, not the raw model. Click a card for its full spec.

## 6 agents Live data · click a specimen for its anatomy

## Field notes

Depth in one industry, not breadth across many

Vertical SaaS is the largest slice of the catalogue at 31 agents, more than four times the frontier tier. These are not general assistants pointed at an industry; they are built around one profession's workflow, vocabulary and system of record.

The moat is domain depth rather than model quality. Knowing which fields a claims adjuster checks, or which clause a reviewer flags, is not something a better model supplies on its own. Expect this tier to keep growing fastest, because every profession with a repetitive document workflow is a candidate and the incumbents already hold the data.

- [Brandon Chaplin](https://www.linkedin.com/in/brandon-chaplin-digital-marketing-strategist)

## Common questions

### What is a vertical AI agent?

An agent built for one industry’s workflow and sold as software: claims handling, clinical coding, contract review and the like. It ships with that industry’s data model, terminology and integrations already built in.

### What is the difference between vertical and horizontal AI?

Horizontal tools work for any industry and leave the domain knowledge to you. Vertical tools do one industry properly, including its rules, file formats and system of record. Vertical products usually run on a general model underneath and add the domain layer on top.

### Why use an industry-specific AI agent instead of ChatGPT?

Because the hard part is rarely the reasoning. It is knowing which fields matter, which exceptions apply and how to write back into the system of record. A general assistant has to be told all of that every time; a vertical product has it built in.

### Will AI agents replace vertical SaaS?

Unlikely on their own. The durable assets in vertical software are the proprietary data, the integrations and the compliance work, none of which a general agent has. The more common pattern is existing vendors adding agents to what they already own.

## Next in the learning path

- [The full catalogue Every agent across all verticals](https://agentfieldbook.org/catalog/all/)

- [Agent platforms The horizontal substrate underneath](https://agentfieldbook.org/catalog/platforms/)

- [Legal agents A worked example of a deep vertical](https://agentfieldbook.org/catalog/legal/)

- [Finance agents Another data-heavy, regulated vertical](https://agentfieldbook.org/catalog/finance/)

## Entries

_6 entries listed on this page._

| name | description | subjectOf |
| --- | --- | --- |
| Salesforce Agentforce | A governed enterprise runtime wrapped around CRM, actions, and flexible model supply. | https://www.salesforce.com/news/press-releases/2026/05/27/fy27-q1-earnings/ |
| Intercom Fin | Rebranded from Intercom to Fin; agreed to be acquired by Salesforce (~$3.6B, Jun 2026). Flagship Fin now runs on proprietary Apex models, not Claude. Strong compliance signalling (ISO 42001). | https://venturebeat.com/technology/intercoms-new-post-trained-fin-apex-1-0-beats-gpt-5-4-and-claude-sonnet-4-6 |
| HubSpot Breeze | Outcome pricing by resolution is strategically notable. | https://www.hubspot.com/company-news/hubspots-customer-agent-and-prospecting-agent-now-you-pay-when-the-task-is-complete |
| Zendesk AI | Category presence strong; model stack needs a deeper documentation crawl. | https://support.zendesk.com/hc/en-us/articles/5352026794010-About-automated-resolutions-for-AI-agents |
| Freshworks Freddy | Promising operational automation; exact model stack not public. | https://www.freshworks.com/freddy-ai/ |
| Ada | Product strength exceeds public proof of model sourcing. | https://www.ada.cx |