At Main Sequence, our investment strategy has consistently centered on transformative technologies that solve some of the planet’s toughest challenges. And there’s no better example of this than at the intersection of AI with the physical world, especially when applied to Australia’s (not to mention the planet’s) most critical industries. From our early investments in sensing and navigation technologies (Baraja, Advanced Navigation), computer vision applications (Lumachain, Presien, Arkeus) and even more complete autonomous solutions (Emesent, RIOS, Swoop Aero), we've been excited about the role advanced digital technologies can have in more traditional industries.

Over nearly 8 years of Main Sequence, we’ve invested over $100m in Physical AI companies, making it one of our most substantial areas of focus.  

Simultaneously, we've backed AI-powered enterprise software that harnesses natural language processing to extract intelligent insights from digital text (Kapiche, Castlepoint). But the generative AI revolution has sparked a more profound question: How can we translate the remarkable reasoning capabilities of large language models into physical intelligence?

We believe we are at an inflection point. As hardware costs plummet and AI capabilities surge, we're witnessing the convergence of generative AI and robotics—a convergence that promises to transcend the limitations of traditional automation. 

Today's robots are precision instruments trapped in rigid frameworks. They excel in controlled environments but struggle when confronted with complexity, context, and nuanced decision-making. Meanwhile, critical industries—security, mining, agriculture, manufacturing, logistics—face an existential labor shortage that no conventional automation can resolve.

The challenge is no longer about building faster or more efficient machines, but about creating adaptive intelligence that can truly "reason" in unpredictable physical environments.

Our investment in this space represents more than technological optimisation. It's a fundamental reimagining of how intelligent systems can augment human capabilities, transforming constraints into opportunities.

This is where our recent investment into Breaker enters the story. Breaker’s AI agents represent the bridge between robotic precision and human adaptability that industries have been waiting for.

Breaker is revolutionising robotics by making autonomous systems smarter, more adaptable and trustworthy. Unlike traditional hardware-focused companies, Breaker has developed AI software that transforms robots from passive tools into proactive teammates. Breaker’s platform allows a single operator to control dozens of coordinated systems with real-time, context-aware decision-making.

Founded in 2023 by Matthew Buffa (ex-Anduril), Michael Irwin (ex-DroneShield) and Vanja Videnovic (ex-Hargrave Technologies), the team combines deep experience in AI, robotics and defence. In our early conversations, all three demonstrated a “Get stuff done” mindset and an insatiable hunger for learning—not just vision, but execution.

When we met Breaker, we knew we'd found something special: a vertical AI solution built on deep understanding of customer pain points and real-world expertise.

‍How do we create robots that can truly “think” on their feet?

Breaker’s innovation is on-device AI agents enabling two-way natural language communication and context awareness. By eliminating reliance on external servers, Breaker ensures robots remain reliable even in disconnected, high-risk environments. 

Its technology allows robots to team with unmanned aerial vehicles (UAVs) and other platforms, providing greater flexibility and front-line safety. In essence: robots become more like teammates, less like a liability. While initially focused on defence and security, Breaker's platform has applications wherever humans work alongside complex robotics and sensor systems—from search and rescue to mining, agriculture, logistics, and manufacturing. By enabling operators to communicate high-level intent rather than technical commands, Breaker eliminates the need for specialised robotics training.

The unique combination of expertise in both robotics and GenAI is what sets Breaker apart. It’s not just solving problems but is reshaping the way we think about what robotics can do. 

During one of our first meetings at a major industry conference, Breaker’s booth was busy with interested customers and partners—a strong signal of market demand. This was reinforced with Breaker’s early traction with defence customers in several different markets. 

Only 18 months into its journey, Breaker has opened its US office in Austin and is already working with industry partners to advance its AI capabilities and autonomous technology, including SensorOps, Airvolute and Australian Institute of Machine Learning (AIML). 

We’ve been proud to lead Breaker’s pre-seed round and join the company as it scales its AI agents which enable robots to act, adapt and collaborate like humans. Breaker is on an exciting growth trajectory—using this funding to triple its team over the next six months and further expand its footprint in the US to meet growing demand. Its mix of technical depth, commercial acumen and relentless drive makes Breaker the kind of company we love to back.

‍