Description
Originally a UCL AI spin out, London-based Hazy was initially incubated by Post Urban Ventures and CyLon cybersecurity accelerator. Our startup began trying to fix the flaws of traditional data redaction and then data anonymisation. We soon discovered anonymised data will always pose a risk to re-identification.
We looked at the opportunity in applying machine learning to synthetic data, which starts with the basis that with no real data there is no real risk. We pivoted our focus completely to AI-generated smart synthetic data and have worked to make it both differentially private and statistically equivalent. We quickly discovered that the organisations that would most benefit from smart synthetic data are highly regulated large enterprises with legacy architecture.
We looked at the opportunity in applying machine learning to synthetic data, which starts with the basis that with no real data there is no real risk. We pivoted our focus completely to AI-generated smart synthetic data and have worked to make it both differentially private and statistically equivalent. We quickly discovered that the organisations that would most benefit from smart synthetic data are highly regulated large enterprises with legacy architecture.
