Robotic companies often need to address simple but confounding issues. The robot generates a lot of data. Even simple robots can easily generate up to one day of data per day, as they continuously capture data from cameras and sensors.
Australia-based Alloy believes it will help with that problem. A startup is building a data infrastructure that helps robotic companies process and organize all the data that robots collect from a variety of sources, including sensors and cameras.
In its core, the alloy encodes and labels the data it collects, allowing users to use natural language to search the data to find bugs and errors. Users can also set rules to catch and flag issues in the future. This is similar to how observability tools flag error flag errors in software code.
“The current pattern is to look for some anomaly and then play the data,” Joe Harris, founder and CEO of Alloy, told TechCrunch. “They’ve since gone on to spend hours scrubbing this data and looking for these issues that have been flagged on them, and trying to diagnose this one-off edge case if it happened before or not, if it actually happened.”
This data problem continues to be complicated when considering the amount of data produced by a single robot, just as robotic companies are aiming to scale, Harris added.
Harris has been fascinated by robotics since he was a child. However, when he graduated from university in 2018, he had few opportunities to work in this field, and instead played multiple roles in Australian high-tech companies, including Atlassian and Telehealth startup Eucalyptus.
In 2024, he decided that starting his own robotics company was the right time. He originally thought that he was focusing on the construction of agricultural industry robots due to his interest in vertical agriculture, but when he began talking to other founders, the issue of managing the product of data robots continued to be maintained. He thought he might solve the problem first.
TechCrunch Events
San Francisco
|
October 27th-29th, 2025
“If you need to solve this problem for yourself and my robotics company, I have a great horizontal solution,” Harris said. “Maybe it will be a more important, short-term mission, so that other robotics companies can spend less time plumbing and more time reaching that high reliability.”
Since its launch in February 2025, Alloy has signed four Australian robotics companies as design partners, and is about to enter the US market this year.
“The customers we found are most excited about this because they have experienced the pain of building and maintaining themselves,” Harris said. “They have, rather, great tools like DataBic, built specifically for robotics.”
Additionally, with participation from AirTree Ventures, Xtal Ventures and Skip Capital, Alloy raised more than $4.5 million (approximately $3 million) in the pre-seeded round led by Blackbird Ventures.
The company still doesn’t have many direct competitors. Many robotic companies are either remodeling existing data management tools that are not designed to prevent multimodal data robots from being generated, or are trying to build their own internal data management tools.
With commercial use cases in robotics continuing to rise, Alloy hopes to gain a significant share of the growing market.
“It wasn’t a time before that we were building a robotics company,” Harris said. “We really want to enable the next 10,000 100,000 robotics companies that don’t exist yet, so they don’t necessarily have to reinvent the wheels as all companies have.”
