Figura Analytics
Figura Analytics is a spinout business from Loughborough University. Its academic founders spent many years developing a patented method for characterising particles within a solution, from beverages through to fluids such as blood and saliva. The company's mission is to transform the speed and accuracy of liquid analysis, replacing methods that take days with a process that delivers results in minutes.
The client
Qubiz is more than Figura Analytics' technology partner. Through the InQubator by Qubiz, we are also an investor and co-creator in their groundbreaking solution for contamination detection, supporting the company from its earliest software foundations through to its readiness for market.
The challenge
Figura Analytics was founded by a team of scientists with deep expertise in nanopore research, but without the in-house software specialists needed to turn that science into a commercial product. They required a partner who could build their applications to a high standard, within budget, and on a realistic timeline.
The technical demands were considerable. The solution needed to interpret very large volumes of sensor data, build a reliable “fingerprint” for each particle, and determine whether a sample contained contaminants, all at a speed that could challenge established laboratory methods. Traditional liquid analysis can take several days; Figura's ambition was to deliver an answer in minutes, which meant the software had to process tens of thousands of data sets every second in real time, without compromising accuracy.
Just as importantly, the product had to be usable. An earlier version of the application was highly technical, crowded with buttons and configurations that demanded expert knowledge to operate. For Figura to gain customers and scale, the software needed to become intuitive enough for any operator to run with confidence.
WHAT WE BUILT
Controls the valves, pumps and sensors of Figura's hardware, processing the liquid and capturing the resulting data. Built from a stripped-back, rebuilt fork of an open-source amplifier application, removing everything that wasn't needed and rebuilding the rest around exactly what Figura required. Written in C++ to deliver the performance real-time processing demands.
Built entirely from scratch by Qubiz, this application interprets the captured data, presents it through clear graphs, and compares each sample against a fingerprint to determine whether it falls within expected parameters.
Working with a dedicated graphic designer, Qubiz replaced a dated, intimidating interface crowded with expert-only controls with a clean, modern and far simpler one. Operators can now configure a sample, start the process and step away — the system clears obstructions automatically, and only raises a banner and audible alert if manual intervention is genuinely required.
Shape analysis characterises the particles detected within each sample. Separately, a Random Forest classifier compares the sample against an expected fingerprint, analysing 50,000 to 60,000 data sets per second in real time and giving the operator a clear visual pass or fail indication. Fingerprints, projects and captured data are shared through the cloud, replacing a manual file-handling process.
Four takeaways from building Figura's software ecosystem end to end.
As an investor and co-creator through the InQubator, Qubiz was able to shape the solution alongside Figura's scientists, ensuring the technology served both the science and the business.
The redesign demonstrated that even the most advanced technology only succeeds when people can use it easily. Simplifying the experience was as valuable as any technical gain.
Choosing C++ and a performance-focused stack made it possible to process tens of thousands of data sets per second in real time, proving that thoughtful technical decisions underpin real-world results.
Owning the full development, from hardware control through to data analysis, allowed Qubiz to optimise resources and deliver a complete, dependable solution within budget.
The outcome
2-3 min
Per measurement run, down from several days with traditional liquid analysis.
50k+/sec
Data sets analysed in real time by the machine learning fingerprinting algorithm.
2
Purpose-built applications delivered: the Figura Data Reader and Figura Data Analyser.
30 min
Full capture-and-analysis cycle, completing comfortably within half an hour.
This software is transformative, capturing the data and making it ready to use the second the user wants to do so. It has had a massive impact on the research and methods we've been able to deploy.
Figura Analytics shows what becomes possible when deep scientific expertise is matched with the right software partner. By combining custom application development, machine learning and a relentless focus on usability, Qubiz helped turn a patented laboratory method into a market-ready product. If you are ready to bring your own vision to life, Qubiz can help you build intelligent, high-performance solutions from the very first concept through to launch and beyond.
Partner with us
Unified sensor, WMS and ERP data into a single operational platform — eliminating manual reporting and enabling predictive maintenance at scale.
Built an end-to-end intake and compliance pipeline that processes millions of transactions per year with zero manual intervention.