# Applied and Vertical AI
> Global category of startups deploying AI into specific industries, from US legal and healthcare to India's sovereign model stack, attracting record venture capital as of 2026.

**Meta:** type: reference · date: 2026-07-03 · heads:  · 4 takes · 4 lenses · 2 regions

## What it is

Applied AI, also called vertical AI, covers startups that deploy large language models and machine-learning systems to solve domain-specific problems inside a single industry, rather than build foundation models or general-purpose infrastructure. The primary market is United States enterprise software, though substantial companies have emerged across Europe, India, and South Korea. The category divides into two types. Companies that own a proprietary data moat, whether molecular structures in drug discovery, patient records in clinical care, or case-outcome databases in law, can build competitive advantages that a general-purpose frontier model cannot replicate. Companies that re-bundle foundation model capability into a managed workflow without unique data face commoditization as underlying models improve. The durable businesses are defined by data type, not industry sector.

## History

The modern vertical AI wave began in earnest in 2023 as GPT-4's longer context window and improved instruction-following made reliable enterprise deployment feasible for the first time. Harvey (United States, legal research and contract drafting) raised its first significant round in 2023 and reached US$200 million in annual recurring revenue by early 2026. Ambience Healthcare (United States) and Abridge (United States) began converting clinical documentation time-savings into hospital procurement contracts in 2024. Chai Discovery (United States, antibody design) grew from a US$150 million to a US$1.3 billion valuation in fifteen months on a proprietary protein-structure model. The geographic scope widened from 2024 onward as sovereign-AI mandates in India, Saudi Arabia, and Singapore drew domestic capital toward locally built vertical applications.

## Current state

As of mid-2026, CB Insights' annual AI 100 report identifies healthcare and financial services as the largest vertical subcategories, with nine companies each among the hundred most promising AI startups globally. Vertical AI companies raised an estimated US$3.5 billion in 2025, triple the prior year, even as the bulk of Q1 2026's record US$242 billion in global AI venture funding went to foundation-model labs. The a16z Big Ideas 2026 analysis finds leading vertical software companies in healthcare and legal reaching US$100 million ARR within a few years of founding. Stanford's 2026 AI Index reports 88% of organisations using AI in at least one business function, though fewer than 10% have fully scaled AI in any single process, pointing to a large gap between pilots and production deployment that vertical specialists are positioned to close.

## Relationships

Most vertical AI companies run on foundation models from a small set of United States providers: OpenAI, Anthropic, Google DeepMind, and Meta's open-weight Llama series. This creates upstream dependency risk, which materialized in June 2026 when the United States government restricted foreign-national access to leading Anthropic models, accelerating domestic model investment in India and elsewhere. [Sarvam's June 2026 unicorn round](/ko/n/sarvam-ai-unicorn-hcltech-2026) (India, US$234 million led by HCLTech at a US$1.5 billion valuation) is the clearest illustration of this reaction: a locally built model stack for Indian languages deployed across banking, insurance, and government services. [TwelveLabs](/ko/n/twelvelabs-series-b-100m-2026) (South Korea-founded, United States-based) shows a different pattern, video-native foundation models targeting enterprise archives, backed by Korea's NAVER and Amazon with a US$100 million Series B in July 2026. France's [Alan](/ko/n/alan-series-g-2026) (health insurance and primary care, a €480 million Series G at €5.5 billion in June 2026) reflects European vertical AI competing on regulatory fit and local data access.

## What to watch

- Model commoditization: as open-weight models close quality gaps with closed frontier models, vertical companies without proprietary data face margin compression and potential displacement by a better general-purpose alternative.
- Regulatory arbitrage: United States export restrictions on leading AI models are accelerating non-US domestic AI investment, creating parallel vertical ecosystems in India, South Korea, and the Gulf states.
- Revenue durability: the central investor question for the second half of 2026 is whether vertical AI ARR reflects durable workflow integration or substitutable pilots that a better foundation model could displace.
- Sovereign procurement: government contracts in healthcare, defense, and public administration are becoming a distinct revenue channel, rewarding companies that meet local data-residency and security requirements.

## Regional takes (batched by bias / lens)

### industry analysis
- **CB Insights AI 100 2026** (Global, en) — CB Insights annual ranking of the 100 most promising AI startups: healthcare and financial services tied as the largest vertical subcategories (9 companies each); the report identifies data-type moats, not industry sector, as the defining trait of durable vertical businesses; drug-discovery firm Chai Discovery grew from a US$150m to a US$1.3bn valuation in 15 months.
  Source: https://www.cbinsights.com/research/report/artificial-intelligence-top-startups-2026/

### official record
- **Stanford HAI 2026 AI Index Report** (Global, en) — Stanford annual AI index: US private AI investment reached US$285.9bn in 2025, more than 23 times China's US$12.4bn; 88% of organisations use AI in at least one business function; fewer than 10% have fully scaled deployment in any single function.
  Source: https://hai.stanford.edu/ai-index/2026-ai-index-report

### venture capital analysis
- **Andreessen Horowitz Big Ideas 2026** (United States, en) — a16z partner perspectives on vertical AI for 2026: top vertical software companies in healthcare, legal, and housing are reaching US$100m ARR within a few years; financial services incumbents will rebuild core infrastructure around multi-agent AI rather than layering tools on legacy systems.
  Source: https://a16z.com/newsletter/big-ideas-2026-part-2/

### venture data
- **Crunchbase News (Q1 2026 global venture record)** (United States, en) — Crunchbase data report: US$300bn raised globally in Q1 2026, an all-time quarterly record; AI companies captured US$242bn, 80% of all venture capital deployed; vertical AI raised an estimated US$3.5bn in 2025, triple the prior year.
  Source: https://news.crunchbase.com/venture/record-breaking-funding-ai-global-q1-2026/

## Across the graph
- Related: [[alan-series-g-2026]], [[twelvelabs-series-b-100m-2026]], [[sarvam-ai-unicorn-hcltech-2026]]
- Entities: Applied AI Startups, Foundation Model Startups, Enterprise Software, Mega Rounds, Region:silicon Valley

---
Canonical: https://rbtfl.xyz/ko/n/applied-ai-startups-dossier