5 Interesting Startup Deals You May Have Missed: Floating Nuclear Power, Robot Report Cards And Voice AI For Farmers

This is a monthly column that runs down five interesting startup funding deals that may have flown under the radar. Check out our previous entry here.

From putting nuclear reactors on barges to grading how well AI models can control robots, this month鈥檚 crop of interesting startup deals takes AI and other emerging technologies well beyond the conventional software stack.

Other companies that caught our eye are applying automation to the decidedly old-school worlds of building-material procurement, commercial property maintenance and farm recordkeeping. Let鈥檚 take a closer look.

$50M to put nuclear power at sea

Nuclear power plants are famously difficult and time-consuming to build. thinks putting them on barges could offer another way.

The Long Beach, California-based startup raised what it says was an oversubscribed $50 million seed round led by , just two months after emerging from stealth with $10 million in pre-seed funding. The of investors in the deal included , , , and others.

Bluecore is developing compact, water-cooled small modular reactors designed to operate aboard floating barges. Rather than spending years constructing a new power plant and the accompanying infrastructure on land, the idea is to manufacture the systems and move them to where electricity is needed.

鈥淥ur focus is simple. Create and deliver zero-emission energy as safely and quickly as possible,鈥� CEO and founder wrote in a social media . 鈥淥ver 3 billion people live within an hour of water. We want to power them all.鈥�

Its first target is the Port of Long Beach, with other ports and power-hungry AI data centers among the potential customers. The company says it鈥檚 working with the and as it pursues certification.

Bluecore鈥檚 raise comes amid a broader nuclear funding boom. Nuclear fission startups alone pulled in roughly $2 billion in venture funding in 2025, per SA国际传媒 data, and investors have continued writing enormous checks this year. More broadly, cleantech-, EV- and sustainability-focused startups raised about $15 billion in the first half of 2026, with second-quarter funding reaching its highest quarterly level since 2024.

Related SA国际传媒 query:

$31M to bring AI to the building-materials business

Ordering cement and steel might not sound like an obvious AI use case, but Saudi Arabia-based sees a lot of room for improvement.

The Riyadh startup last week said it had secured $31 million in new capital, consisting of a $13 million Series B equity round co-led by , 鈥檚 venture arm, and , plus an $18 million growth-debt commitment from under a previously announced facility. The company has now raised more than $83 million, .

BRKZ operates a marketplace that connects construction companies with suppliers of building materials, while also handling sourcing, logistics and financing. More interestingly, it says it has amassed some 38 million structured data points that power an AI pricing engine trained on roughly 40,000 requests for quotes.

The company says 84% to 89% of its predicted prices come within 5% of the eventual transaction price. Another AI agent reads photos of cement delivery notes sent through , matches them to orders and verifies deliveries 鈥� with roughly three-quarters processed without human intervention.

BRKZ is riding a in Saudi Arabia even as startup investors remain selective about construction and property technology more broadly. Global proptech startups raised about $6.5 billion via roughly 640 deals in the first half 2026, SA国际传媒 data shows. That’s on pace to top last year鈥檚 dollar figures, even as deal count has dipped this year. Similar to other startup sectors, investors in proptech are increasingly directing capital toward companies using AI and automation to cut costs and streamline operations.

Related SA国际传媒 query:

$24M for robots to take care of commercial properties

Robots are already assembling cars and moving packages around warehouses. wants them to also mow lawns and sweep parking lots 鈥斕齛nd do security patrols of those properties while they鈥檙e at it.

The Santa Clara, California-based autonomous robotics startup last week said it has raised a $24 million Series A led by to scale its fleet of robots designed to handle outdoor property maintenance for commercial real estate owners and operators.

Its machines combine autonomous navigation with attachments that allow them to perform jobs like sweeping, debris removal and landscaping tasks 鈥� all while conducting 鈥渟oft security鈥� 鈥� across large campuses and commercial properties.

Rather than trying to build a general-purpose humanoid robot, Viabot is applying autonomy to repetitive jobs that property owners already pay people and contractors to perform. The startup, which operates on a 鈥渞obot as a service model,鈥� sees an opportunity to fill a labor shortage for what鈥檚 often considered 鈥渄irty, dull and dangerous鈥� outdoor work, , founding partner at , told SA国际传媒 News in 2021, when the company raised earlier funding.

Startup investors are pouring money into those sorts of real-world AI applications. Global venture funding to physical AI companies 鈥� including robotics, autonomous vehicles, aerospace, drones, industrial automation and sensors 鈥� reached $47.4 billion across 521 deals in the first half of 2026, SA国际传媒 data shows. That’s nearly 4x the $12 billion invested in the second half of 2025 and almost 80% above the year-ago period.

Related SA国际传媒 query:

$10M to give AI robots an independent report card

As frontier AI models move from controlling software to controlling robots and machines in the real world, wants the public to have an independent way to understand what they can actually do.

The 3-month-old San Francisco startup announced last week that it raised a $10 million seed round led by , with participation from , , and others.

Robocurve says it wants to act as an independent third-party auditor for physical AI, testing how well frontier models can control real robots and publicly reporting the results. The company says its own research has already found that general-purpose large language models can outperform specialized robotics vision-language-action models on some simple tasks.

Importantly, Robocurve isn’t positioning itself as a conventional robotics benchmarking startup. It is incorporated as a Public Benefit Corporation with a legal duty to independently evaluate the robotics capabilities of frontier AI systems and report those findings to the public.

The company says AI labs don’t dictate its research agenda, evaluation methodology or published results, and it plans to work with governments, policymakers and civil society as robot capabilities advance.

Academia is a big part of that model, too. Rather than developing every benchmark itself, Robocurve funds academic teams and supplies them with robot hardware to create open-source benchmarks. More than 200 institutions 鈥� including researchers from 19 of the world’s top 20 universities, according to the company 鈥� have signed up for its benchmarking program. Robocurve is offering a combined $500,000 in funding plus free robotic arms to participating academic groups.

Its funding is timely given the massive influx of capital pouring into robotics and physical AI. Within the broader physical AI sector, robotics startups alone raised more than $21 billion globally in the first half of 2026, SA国际传媒 data shows, already eclipsing the nearly $16 billion raised in all of 2025 and even the $15.3 billion invested during the venture market’s 2021 peak.

As ever more powerful models move from screens into machines capable of manipulating the physical world, figuring out what those models can 鈥� and can’t 鈥� safely do becomes a more consequential problem.

Related SA国际传媒 query:

$5M to let farmers talk instead of type

A lot of agricultural software has one basic problem: Farmers don鈥檛 spend their days sitting at desks.

With that in mind, this month raised a $5 million pre-seed round led by for a voice-first AI platform designed specifically for farmers, agronomists and other agricultural workers.

Instead of asking someone working in a vineyard or almond orchard to stop and fill out a form or type notes into a computer, Tellia enables them to leave a voice note, send a message, or even submit a photo to log records, generate reports and set reminders. It says its AI then turns that unstructured information into records associated with the correct field, crop and crew.

For example, for a livestock farmer that might mean a voice prompt like: 鈥淭ellia, the vet just checked Herd 3. All clear, next health check due in 6 weeks, log that.鈥�

Or a vineyard manager might ask: 鈥淭ellia, based on this year’s Brix and pH logs, what’s the projected alcohol level for the Cabernet lot?鈥� and receive an instant answer based on previously collected data.

San Francisco- and Paris-based Tellia was founded last year and says its technology is already deployed across 1 million acres, including at and wineries in the U.S., as well as agricultural organizations in Europe.

and also participated in its latest funding.

Its raise comes amid a much tougher environment for agtech startups overall. Venture investment in agriculture and farming remains in a correction from its 2021 peak, when startups in the space raised $10.5 billion across more than 1,400 deals, SA国际传媒 data shows. That makes companies applying increasingly cheap and accessible AI to specific, everyday farming problems an interesting corner to watch. Voice AI, in particular, has emerged as one of the hot spots in artificial intelligence funding in recent years.

Related SA国际传媒 query:听

SA国际传媒