When CrossTech launched, it needed scalable cloud infrastructure for its advanced AI platform. However, its initial experience with a third-party cloud provider proved challenging. CrossTech Founder and Managing Director Haydon Bartlett-Tasker found the platform "clunky, complex, and painful. We were spending lots of time managing the platform instead of building our product." The complicated interface of the third-party platform also made it hard to configure correctly, leading to unexpected cloud costs. 

As existing users of Google Workspace, CrossTech’s search for a more intuitive, efficient cloud provider naturally led it to Google Cloud. CrossTech was impressed with the easy-to-use interface and responsive support of Google Cloud, and recognized that its high-performance architecture, including Compute Engine and Cloud Run, could deliver the real-time data processing and containerized microservices it needed. Importantly, CrossTech saw Google Cloud as the perfect home for its AI platform. 

"As an AI company, we wanted a cloud provider that could provide the performance we needed as we scaled," Bartlett-Tasker explains. "We saw that Google Cloud was making large investments in its own compute infrastructure, so we knew it would outperform our existing provider."

CrossTech migrated to Google Cloud in 2020 and built a sophisticated containerized microservices architecture to power its computer vision pipeline. With strict customer service level agreements requiring them to process vast amounts of inspection data in very short periods, CrossTech needs immense computational power. CrossTech’s microservices architecture, leveraging Cloud Run and the auto-scaling capabilities of Compute Engine, automatically scales its resource usage up and down to meet demand, meaning it only pays for what it uses.

As a computer vision company, the core intelligence of CrossTech's products lies in its AI models. With the integrated, streamlined environment of Vertex AI, the team has been able to significantly speed up its model development lifecycle, including using Vertex AI Notebooks for rapid prototyping and testing. 

"With Vertex AI, the infrastructure is taken care of, allowing us to focus on writing our custom model training or building on existing models," Bartlett-Tasker explains. "Thanks to Vertex AI, we’ve halved the time it takes to build the first version of a model to just six weeks. That means we can take features to market sooner. If they don’t work, we save on opportunity cost. If they're successful, we can capture the first-mover advantage."

In the competitive AI space, speed is paramount, and with the integrated tools and scalability of Google Cloud, CrossTech has not only sped up development, but deployment too. This is thanks to the company’s use of App Engine, which, as a fully managed service, automates much of the work that typically slows down deployments, helping CrossTech to reduce deployment cycles by 70%. This improves team morale by avoiding frustrating deployment headaches, while enabling CrossTech to deliver innovative features to customers more quickly.

As it grows, CrossTech continues to look for ways to increase efficiency. It’s currently experimenting with Google Agentspace to automate tasks and streamline internal agentic workflows, allowing it to remain competitive and agile. It’s also bringing more generative AI services, such as Firebase Studio, into its software development process, both to simplify and streamline coding and to attract emerging developers who are growing accustomed to using gen AI tools for problem solving and code generation. 

Firmly established in the UK, CrossTech is now turning its focus overseas. This raises new challenges, such as navigating diverse data sovereignty laws and supporting a larger, geographically dispersed workforce. With Google Cloud, the company is confident it can easily scale its operations and minimize administrative overhead, with its global infrastructure and regional data centers helping CrossTech to provide a reliable service while complying with data sovereignty requirements. Ongoing access to Google Cloud experts, meanwhile, will be a valuable resource for CrossTech's growing teams. 

For Bartlett-Tasker, it’s this ability to scale with Google Cloud that has helped get CrossTech where it is today. "With Google Cloud, we’ve built one of the UK’s most advanced computer vision-based systems," says Bartlett-Tasker. "Google Cloud is powering that national capability, keeping trains running on time, keeping infrastructure safe, and helping get passengers where they need to go. None of that would have been possible without Google Cloud."