Sage unveils data-driven musical AI transparency demo
How the composition was created
The project was built with Loud Numbers, a specialist in data sonification, and performed by a live string quartet. Changes in financial metrics are mapped to pitch, tempo and volume.
By converting numbers into sound, the piece demonstrates Sage’s “Glass Box” philosophy, which calls for AI outputs that are inspectable, traceable and explainable rather than hidden inside a black box.
Why explainability matters to finance teams
Finance departments are increasingly using AI for analysis, workflow automation and operational tasks. Sage’s own research shows 71% of finance leaders reject AI-driven decisions they cannot explain.
The same study found finance professionals spend more than 12 hours each week rebuilding logic, validating figures and defending AI-generated conclusions.
Financial Symphony draws on anonymized customer datasets from the UK, US, France and South Africa, ensuring the music reflects real-world business activity.
Rather than using data as vague inspiration, the team applied strict rules so each audible shift directly corresponds to a financial change.
Details of the musical layers
The string-quartet layer uses UK accounting data spanning Q1 2015 to Q1 2026. The length of each note mirrors average invoice-payment speeds during that period.
An electronic and percussive layer incorporates Q1 2026 data from the US, France and South Africa, representing the continuous pulse of global trade.
“Businesses can’t build confidence on answers they can’t trust,” says Aaron Harris, Chief Technology Officer at Sage. “Our Glass Box approach means that our customers will always know how Sage’s AI got to the answer.”
Harris added that “Every note, rhythm and change in volume has its origins in real business data and the journey from data to music is transparent.”
Related Post: Revolut Business hits £4.5bn revenue with customer-focused banking innovations
The project highlights a core governance question: can users trace an AI output back to the data and assumptions that produced it?
Beyond the artistic angle, Sage positions Glass Box as an alternative to opaque AI systems that deliver recommendations without exposing reasoning.
In a finance context, that distinction affects auditability, exception management, internal controls and the ability to explain decisions to CFOs, regulators or external auditors.
Key practical requirements include traceable data sources, visible assumptions, interrogable outputs, point-of-decision explanations, human accountability and auditable financial operations.
By assigning each sound an identifiable link to underlying data, Financial Symphony makes those requirements tangible for users.
Finance leaders are likely to evaluate AI systems not only on speed and accuracy but also on whether results can be challenged, governed and defended.
For Sage, explainability is becoming both a product feature and a trust factor; the company demonstrates that in high-stakes financial operations, the most valuable AI is intelligible rather than just intelligent.
Loud Numbers analyzed the anonymized datasets and defined the rules that translate financial signals into musical elements, drawing on its experience across classical, techno and sound-art genres.
In April 2026, Sage and PwC launched the “Beyond the Black Box” initiative at the Sage Future event, aiming to make explainable AI practical rather than purely theoretical.
A concurrent agreement with AWS brings cloud reliability, flexibility and security expertise to Sage’s AI-powered offerings for small and mid-sized businesses.
