Way back in 2011 Marc Andressen, Silicon Valley venture capitalist, told us that “software is eating the world”. Quite far back, in 2017, Jensen Huang, CEO of Nvidia, told us that “Software is eating the world, but AI is eating software”, but not many people were listening.
When ChatGPT launched at the end of November in 2022 it was only a few hours before users were reporting its ability to write code. The industry finally heard the message and has been talking about AI eating software ever since, and how its appetite has grown from “fancy auto-complete” to “software dark factory” over the last 4 years.
While AI threatens to replace all things code, the reality is that all things are not code, even if they rely heavily on code to be accomplished. This is the “when you only have a hammer, every problem looks like a nail” in fancy clothes. Not every startup, nor every SaaS, is going to be replaced by a guy in marketing vibe-coding on the weekend.
Some businesses have moats that even a factory of agents can’t build you a bridge to cross. In this article we are going to look at a group of Australian startups, some founded pre-ChatGPT moment, some post-, that have landed successful fundraising rounds recently in part because they do have a moat.
Breaker: Coordinating military robots
Breaker was founded in 2023, making it one of the companies in this group created after ChatGPT. It develops software that allows military operators to control and coordinate multiple autonomous systems rather than operating each drone or robot separately.
Its software is designed to work across different robotic platforms operating in the air, on land and at sea. One of its interfaces allows an operator to give instructions by voice, with Breaker’s agents translating those instructions into actions across the autonomous systems.
The company is not manufacturing the underlying robots, but its product still has relationship-driven and physical-world moats. It’s built up the networks, and it has done the hard work of integrating with different hardware, communications systems and military operating environments.
Breaker raised A$9 million in seed funding in February 2026. The round was led by Bessemer Venture Partners, with Main Sequence making a follow-on investment after leading Breaker’s pre-seed round less than a year earlier.
Brumby: Autonomous drones for moving cattle
Brumby, originally GrazeMate, was founded by Sam Rogers in 2025. The company develops autonomous drones that can muster cattle without requiring a farmer to pilot the drone manually.
A farmer selects a paddock and a task through an app, after which the system uses AI to react to the movement of the herd and move cattle towards the required destination. The company has been developing reinforcement-learning models intended to reproduce low-stress stockmanship techniques. The same drones can also collect information about cattle, pasture, water and infrastructure while operating across a property.
Brumby is building its moat on field data and operating experience from autonomous cattle mustering. Each deployment adds information about animal behaviour, terrain and farm conditions, while improving a system that has to work reliably across different properties and herds.
The company raised A$1.2 million in pre-seed funding in early 2026, with the round led by Y Combinator with backing from Antler and NextGen Ventures.
Superstat: Computer vision for amateur sports
Superstat uses computer vision to extract statistics from video of amateur sport. A basketball game, for example, can be filmed on an ordinary phone and uploaded to Superstat, which identifies players, tracks the ball and produces statistics covering shots, passes, turnovers and other parts of the game.
That makes Superstat an interesting example of a company that does not need to manufacture its own hardware to develop a real-world moat. Its model has been developed over 18 months and retrains as more games are uploaded. If the company succeeds in accumulating years of video and performance histories for players, its underlying dataset becomes difficult for a newly generated software competitor to recreate.
The company is initially concentrating on basketball, particularly the much larger US youth and college markets. It says it has already passed A$500,000 in annual revenue.
In July 2026, Superstat raised A$3.5 million in pre-seed funding, led by Blackbird Ventures after taking part in Startmate’s Summer 2026 accelerator. Other backers named in the round included Blinq co-founder Jarrod Webb and Mindset Health founders Alex and Chris Naoumidis.
Yes. The sharper description is GPS-independent navigation, with the moat in specialised hardware and deployment where GPS is unreliable or denied.
Advanced Navigation — navigation systems that operate without GPS
Advanced Navigation’s moat is its specialised navigation hardware and sensor-fusion technology. Its systems are designed to maintain accurate positioning when GPS is unavailable, unreliable or deliberately jammed, an increasingly important capability in defence, aviation, marine and other critical systems.
Founded in 2012 by Chris Shaw and Xavier Orr, the company develops inertial navigation, photonic sensing and positioning systems for use across air, land, sea and space.
In March 2026, Advanced Navigation raised US 158 million, in a Series C round. AirTree Ventures led the round, with Quadrant Private Equity and the National Reconstruction Fund Corporation participating.
Emesent: Autonomous mapping for mines and difficult environments
Brisbane-based Emesent was founded in 2018 and develops autonomous mapping systems for environments where GPS and conventional communications may not work. Its best-known product combines its Hovermap lidar unit with drones and other robotic platforms to map underground mines, tunnels and other difficult or hazardous locations.
The company has expanded from mapping into autonomy. Its Cortex platform lets machines navigate and perform tasks in complex environments, while Aura provides cloud software for processing and managing the resulting spatial data.
Emesent sells into mining, defence and construction, giving it a combination of specialised hardware, autonomy software, field data and integrations with customers operating in environments that are difficult to reproduce in a lab.
In July 2026, Emesent announced A$25 million in new funding. That consisted of a A15 million equity round backed by Main Sequence, QIC Ventures, Orion Resource Partners, Hostplus and NGS Super.
Arkeus: AI sensing systems for defence platforms
Arkeus was founded in 2020 and builds sensing systems for drones and other autonomous defence platforms. Its technology combines specialised optical and hyperspectral sensors with AI that processes sensor data at the edge, allowing a platform to detect and identify objects in conditions where conventional cameras can struggle.
The company’s systems can operate across a wider section of the electromagnetic spectrum than a normal camera, including ultraviolet and infrared. Arkeus has integrated its technology with several drone platforms and has won work with Australian and US defence customers.
Its moat is therefore not simply the AI model. It includes the sensors feeding the model, manufacturing, defence integration, field testing and procurement relationships.
Arkeus raised A$25 million in a Series A round in May 2026. QIC Ventures led the round, with new investors R+ VC, Folklore Ventures and DYNE Ventures joining existing investors Main Sequence, Salus Ventures and Beaten Zone Venture Partners.
Octopusbot: AI forecasts for crop production
Sydney-based Octopusbot was founded around 2020 by Carolina Ferreira and Rodrigo Cortes. It develops software for forecasting crop production, supply risks and commodity-market conditions.
The platform combines satellite imagery, weather information, historical crop yields, planted areas and commodity-market data. It runs more than 10,000 interconnected AI models and says it can produce crop estimates up to seven months ahead of official and trade forecasts. Its customers include growers, traders, brokers and other businesses in the agricultural supply chain.
Octopusbot is another example where the defensibility is less about owning a foundation model than owning a specialised prediction system and the data needed to improve it. Agricultural models also have to be tested over actual planting and harvest cycles, meaning a competitor cannot immediately manufacture years of validation by building a similar product.
In July 2026, Octopusbot announced its first external funding round, raising more than A$1 million from Folklore Ventures and No Brand.
ProcurePro: Procurement software for construction companies
ProcurePro was founded in Brisbane in 2020 and builds procurement software for commercial construction. Its platform manages the process through which builders scope work, compare subcontractors, negotiate terms and award contracts.
Construction procurement is software, but it is tied closely to a complicated real-world process involving contractors, subcontractors, contracts, budgets and live projects. ProcurePro says its software has been used across more than 6,000 construction projects representing over A$130 billion in project value.
That history, along with integrations into contractors’ operating processes, gives ProcurePro a different type of moat from a generic SaaS application. Its AI can work over structured procurement data from thousands of projects to help builders scope packages, compare subcontractors, identify pricing or contracting issues, and support purchasing decisions with context drawn from previous jobs.
ProcurePro raised A$15 million in a Series B round in May 2026, valuing the company at more than A$100 million. QIC Ventures led the round. Brighter Super participated, along with construction group Bouygues through its venture partner ISAI and existing investors AirTree and Glitch Capital.
How are you building your moat?
There is nothing more 2026 than lots of mentions of AI and drones, often together. Though the drones are often (ignoring defence and Brumby’s mustering) just platforms for sensors collecting the data to feed the AI.
Every startup and every SaaS does not need AI and drones, but like the 2010s-2020s SaaS era saw industry spreadsheets sitting on hard drives being moved into team-based web apps living in the cloud, we’re in a new era of moving complex data into the digital realm because the world now has a tool for working with it (AI of course).
Where are you building your moat? On data? Relationships? Hardware? The complexity of regulations and compliance? The complexity of the real word?
If you ever want to chat about building moats and teams to build them, get in touch.
