5 startup deals: DNA AI, plumber dispatch, tunnel warfare

Key Takeaways

• Radical Numerics raised $50M to build AI that reads and generates DNA sequences, with backing from Stripe co-founder Patrick Collison
• Probook secured $40M from a16z and Sequoia for AI that dispatches plumbers and HVAC technicians
• Defense tech startup Traysar emerged with $25M to develop underground warfare capabilities

Five startup funding rounds from the past month reveal where venture capital is actually going in 2025: DNA-reading AI, software that dispatches your plumber, underground military tech, and fixes for Wall Street's paperwork problem. None of these grabbed front-page headlines, but each represents a distinct bet on where automation can crack open stagnant industries.

Radical Numerics: $50M to read and write DNA

San Francisco-based Radical Numerics emerged from stealth with a $50 million seed round led by Emergence Capital. Obvious Ventures, Triatomic Capital, Factory, and First Spark Ventures also participated. The company says it received pre-seed backing from Stripe co-founder Patrick Collison.

The founding team previously built Evo, one of the first AI models capable of reading and generating DNA sequences at scale. Their new mission is broader: building what they call "general biological intelligence," multimodal AI models that can reason across DNA, RNA, proteins, and other biological data.

The company previewed Omnii, its next-generation genome language model, alongside the funding announcement. Applications span drug discovery, cancer diagnostics, and biosecurity.

“Evo showed that AI can generate DNA and whole genomes, the next generation of models will go further with the ability to control function, and eventually, create entirely new forms of life. Our multimodal models are already far more capable, and we understand the responsibility that comes with that.”

— Eric Nguyen, CEO of Radical Numerics

The dual focus on advancing biological AI while defending against its misuse reflects investor anxiety about frontier models. As AI becomes capable of designing biological systems, the same technology that could cure disease might lower barriers to creating harmful biology. Radical Numerics is positioning itself on both sides of that equation.

Probook: $40M to automate home services dispatch

New York-based Probook raised $40 million across two rounds: a $34 million Series A led by Andreessen Horowitz and a $6 million seed round led by Sequoia Capital. Sequoia participated in both.

The startup is building what it calls an AI operating system for home service businesses, targeting plumbers, electricians, and HVAC contractors. Rather than another chatbot, Probook focuses on dispatch, the coordination layer that determines how quickly a technician reaches your door.

The platform ties together customer intake, scheduling, messaging, and outbound communications. The pitch: technicians spend less time waiting between jobs, office staff spend less time on the phone coordinating them.

CEO George Eliadis built the product from firsthand experience. "I grew up pressure washing in upstate New York with my dad. Six summers in the truck. I spent two to three hours of my day driving between jobs. I'd be up on a ladder washing a house and miss calls because I couldn't hear my phone ringing," he said in a statement.

Probook represents a growing category of vertical AI startups targeting industries that still run on phone calls, clipboards, and spreadsheets. Plumbing, electrical, and HVAC collectively represent millions of small businesses with thin margins and high coordination costs. Automation here is not glamorous, but the market is enormous.

Traysar: $25M for underground warfare

Austin-based Traysar emerged from stealth with a $25 million seed round. While defense investors have poured billions into drones, missiles, autonomous tanks, and military vessels, Traysar is betting the next battlefield is below ground.

The company is developing technology for subterranean combat, tunnel detection, underground navigation, and related capabilities. The Israeli-Hamas conflict highlighted the strategic importance of tunnel networks, and militaries worldwide are grappling with threats they cannot easily see or access.

Details on Traysar's specific technology remain sparse, which is typical for defense startups operating in sensitive domains. But the funding size signals serious investor conviction that underground warfare is an underserved niche in the defense tech boom.

Private markets paperwork: lessons from 1968

One of the month's deals targets a problem with historical resonance: the paperwork crisis in private markets. In the late 1960s, Wall Street trading volume overwhelmed manual processing systems so badly that exchanges closed on Wednesdays just to catch up. That crisis ultimately forced the creation of electronic trading infrastructure.

A new startup argues that private markets, which now handle tens of billions in transactions, are running on similarly outdated infrastructure. The company is building software to modernize cap table management, fund administration, and the cascade of documents that accompany private market deals.

For founders, the pitch is straightforward: less time buried in paperwork, fewer errors in equity tracking, faster closes on funding rounds. For investors, it is the promise of reduced friction across their portfolio operations.

What these deals reveal about 2025 investing

These five rounds share a common thread: they target unsexy problems with large addressable markets. DNA sequencing infrastructure. Plumber dispatch. Tunnel detection. Fund paperwork. None will dominate tech Twitter. All represent billions in potential value.

The investor roster also tells a story. Andreessen Horowitz and Sequoia both backing Probook suggests top-tier firms see vertical AI for trades as a legitimate category, not a niche bet. Emergence Capital leading a $50 million seed for biological AI signals conviction that frontier science startups can command mega-rounds early.

For founders, the lesson is that investor appetite remains strong for automation plays in industries that have resisted software adoption. The bar is not "interesting technology" but "specific problem, large market, defensible position."

Frequently Asked Questions

What is general biological intelligence?

A term coined by Radical Numerics to describe multimodal AI models that can reason across DNA, RNA, proteins, and other biological data, rather than being trained on a single modality like text or images.

Why are investors funding underground warfare startups?

Recent conflicts have demonstrated the strategic importance of tunnel networks. Militaries need better detection, navigation, and combat capabilities for subterranean environments, creating a gap that startups like Traysar aim to fill.

What was the 1960s Wall Street paperwork crisis?

In the late 1960s, trading volume overwhelmed manual processing systems, forcing stock exchanges to close on Wednesdays to catch up on paperwork. The crisis ultimately led to the creation of electronic trading infrastructure.

How much did Probook raise in total?

Probook raised $40 million total: a $6 million seed round led by Sequoia Capital and a $34 million Series A led by Andreessen Horowitz.

Who invested in Radical Numerics?

Emergence Capital led the $50 million seed round, with participation from Obvious Ventures, Triatomic Capital, Factory, and First Spark Ventures. Stripe co-founder Patrick Collison provided pre-seed backing.

Source: Crunchbase News / Marlize van Romburgh