As AI continues to reshape industries, one challenge persists: how do you take AI models from concept to reality in a way that’s scalable, secure and sustainable? For Fuzzy Labs, this is more than just a technical problem – it’s their mission.
The Manchester-based startup was founded by engineers with software and DevOps backgrounds who wanted to bring strong engineering principles into the emerging world of machine learning.
We worked with Fuzzy Labs as part of LASR Validate, the UK’s first programme dedicated to accelerating the development of next-generation AI security products. Over six weeks, Fuzzy Labs and fellow startups tackled real-world challenges across key sectors – including finance, telecoms, defence and national security – gaining expert insight and collaborating to shape the future of secure AI.
With this in mind, we caught up with the company’s co-founder and CTO Matt Squire to hear about their vision, programme experience and how they’re doing things differently. “We started Fuzzy Labs to help people get AI into production,” says Matt. “We’ve built a reputation as specialists in machine learning operations (MLOps), using open-source tools to solve our customers’ biggest challenges in production-scale AI.”
The team’s focus isn’t just on delivering technology but on building trusted partnerships. “The most rewarding part of our journey so far has been building close, collaborative relationships with our customers and seeing the results of that effort,” he notes.
Collaboration and open source at the core
Fuzzy Labs prides itself on doing things differently and two elements stand out for Matt. “Firstly, our collaborative approach is key,” he says. “We work directly with data science and technology teams to understand their unique challenges and implement solutions together. Second is our open-source approach.”
By using open-source tooling throughout their work, the Fuzzy Labs team ensures their clients aren’t left with a black box. “We give customers something which they genuinely own at the end of the engagement,” he explains. “Moreover, we don’t just deliver technology, we also engage in knowledge-sharing activities which ensure that our customers are upskilled in MLOps tools and best practices.” For Fuzzy Labs, it’s not just about building systems – it’s about building confidence, capability and long-term success.
A security-first mindset
Security is fundamental to everything Fuzzy Labs delivers – not an afterthought. “We look at security from an MLOps-first perspective,” says Matt. “Any production-scale AI application needs to consider security at every step, from data and model assets, through to model serving and monitoring infrastructure.”
By working closely with clients to understand their threat models, the team builds solutions that balance innovation with trust and resilience.
The LASR Validate experience
Joining LASR Validate gave Fuzzy Labs the opportunity to engage with new ideas and apply their thinking to new contexts. “Taking part in Validate allowed us to get exposure to problem spaces that we wouldn’t otherwise have access to and that in turn means we can better understand how to integrate cyber security concerns into MLOps,” Matt reflects.
It also provided inspiration from the broader cohort. “One standout moment was hearing lightning pitches from all of the cohort and getting to see first-hand the homegrown ingenuity that makes up the UK’s growing AI security sector.”
Evolving with the industry
As AI evolves, so does Fuzzy Labs. The team is tackling emerging challenges like MLOps for agentic AI, delivering groundbreaking R&D projects to further the subject for all practitioners. Just as importantly, they’re also focused inward, building a workplace that’s genuinely inclusive. “We’re proud of the ethnic diversity in our team,” says Matt, “but as we grow, we’d like to improve representation further, making this a place that’s accessible and supportive for everyone who has both talent and passion for MLOps.”
Building for impact
Now three years in, Fuzzy Labs continues to grow – in capability, community and purpose. As AI adoption accelerates across industries, the team is focused on meeting that momentum with strong, secure and scalable foundations. “As the demand for AI grows, we see an opportunity to continue evolving how we productionise, scale and secure AI through great MLOps,” says Matt.
By championing open source, empowering their customers and embedding cyber security into the heart of AI operations, they’re laying the groundwork for a more secure, transparent and equitable future.
Fuzzy Labs is a standout example of what LASR Validate was designed to do – support ambitious innovators tackling the toughest challenges in AI security with collaboration, creativity and purpose.
Stay tuned for more profiles on AI security innovators from our ecosystem, which you can be part of – just visit the Laboratory for AI Security page for more information.
