Last month, $189 billion in venture capital went to startups globally.

Eighty-three percent of it went to three companies — OpenAI, Anthropic, and one other. The coasts are in a capital arms race over who builds the biggest foundation model. Good for them.

From where I sit in Atlanta, that war creates the conditions for a different kind of winner — and the South is full of them.

Last week, we scored two big victories on the venture capital front for Valor.

We closed new up rounds in two portfolio companies: FirmPilot, which raised a $22 million Series A-1 led by DeepWork Capital with participation from Data Point Capital and Thomson Reuters Ventures, and Arpio, our Durham-based cloud disaster recovery platform that simultaneously launched Azure support — expanding from AWS-only to true multi-cloud. Two very different companies. Same thesis: applied AI built for enterprise buyers who need their data to stay private, their workflows to stay compliant, and their infrastructure to actually work when it matters.

FirmPilot is building the most compelling AI marketing platform in legal services — one that helps law firms acquire clients through AI-generated content tightly integrated with case management systems. Founder Jake Soffer had acquisition interest and chose to raise instead.

Law firm technology spending grew 9.7% in 2025, likely the fastest real growth ever recorded in the industry. Thomson Reuters hit one million CoCounsel users, is spending $200 million a year on productized AI, and just acquired Noetica, an AI-native legal analytics startup. Harvey raised $760 million last year at an $8 billion valuation. When the largest legal information company on the planet, Thomson, puts venture capital behind your growth round, it validates the momentum. FirmPilot is now building a proprietary AI model ten times larger than its current one, with a Q3 2026 rollout.

Cloud Resilience Is the AI Infrastructure Play Nobody's Talking About

Arpio has been quietly dominating cloud-native disaster recovery for AWS — the highest-rated DRaaS product on G2. Last week they expanded to Azure, becoming the only DR platform built from the ground up for cloud workloads across both hyperscalers. For enterprises running complex, interdependent cloud environments — and that's every Fortune 500 — Arpio solves a problem the legacy vendors can't touch. The new round fuels go-to-market at exactly the right time: after the February Azure outage, every CIO in America is rethinking their resilience strategy.

The South has a structural advantage in applied AI that no amount of San Francisco capital can replicate.

Enterprises are increasingly wary of handing proprietary data to general-purpose LLMs — and that wariness is creating a market. The applied AI companies winning enterprise contracts right now run vertical models trained on domain-specific data, deployed inside the customer's own environment. Private pools built on top of foundation models, not replacements for them. The South's density of B2B enterprise headquarters means the companies building these solutions and the companies buying them often share a zip code. Geography is doing real work here, and it compounds over time.

The Southern region has the most Fortune 500 headquarters of any region in the United States. Texas alone is home to 54. Atlanta, Charlotte, Houston, Dallas — these are cities where industrial enterprises, logistics companies, financial institutions, and healthcare systems are headquartered. These are the buyers of applied AI. They need vertical solutions that work inside their compliance frameworks, their procurement cycles, and their data governance requirements.

**Early-Stage AI M&A Is the Exit Story Venture Isn't Pricing In (Because Shh!--Most Profitable Exits for LPs Don't Start With a B)**

EY-Parthenon reports that corporates are snapping up seed and Series A AI startups at prices that once required a growth round — teams of fewer than 100 landing $100 million-plus exits. We received another LOI on one of our AI companies this week. As I heard loud and clear from the audience last week at the CIO 100 where I spoke, enterprises don't want to build applied AI internally. They want to buy it from companies already embedded in their workflows.

Where Southern VC and Applied AI Converge

The foundation model wars are a spectator sport for most of us. The real returns in AI venture will come from applied, vertical-specific platforms where the moat is domain data and workflow integration — not compute. Acquirers are buying applied AI teams before they hit scale because the alternative — building internally — is slower and more expensive.

The South's concentration of enterprise buyers creates a natural proving ground for these companies that doesn't exist on the West Coast, where the customer base skews tech-on-tech.

We've been investing against this thesis for a decade. This week was a good one.

If you're a founder building breakout B2B AI — vertical, applied, enterprise-ready — let our AI analyst Vic give your deck feedback right away--drop it here and let's talk.

PHOTO CREDIT: Valor Ventures, CEO Jake Soffer speaking about FirmPilot's pace setting growth at Valor's annual meeting.

Forget the Foundation Model Wars. The South Is Winning the Applied AI Race.