Company: OmniIndex
CEO: Simon Bain, Ken Hawkins, Marina Simonians
Website: https://www.omniindex.io/
About OmniIndex
Established in 2020, OmniIndex’s mantra of ‘never decrypt’ sets them apart with their patented technology ensuring data can remain encrypted at all times – even while it is being searched and analyzed. They have recently applied this technology to financial fraud to enable actionable AI insights and real-time alerting of suspicious activity from fully encrypted financial transactions and access logs without any risk of exposing this business critical data.
While there are a number of powerful fraud detection tools on the market, OmniIndex’s homomorphic encryption uniquely permits the analysis of sensitive financial data without exposing it. This means regulated information can be included in a single workflow without requiring siloed environments. Consequently, real-time fraud intelligence is generated from all encrypted financial data, with only the actionable results being readable, eliminating the issue of incomplete reports as well as the risk of exposing this critical information.
OmniIndex’s move into FinTech happened organically when an existing Life Science customer began using their Web3 data platform PostgresBC for real-time fraud detection. Health Innovation Hub (HIH) did this after realizing that the same reasons why they could analyze their highly regulated health data with OmniIndex’s Web3 and AI solution also meant they could analyze their encrypted financial data to detect any suspicious activity. As such, they migrated their siloed financial data to their main workflow, storing it in their own OmniIndex blockchain, and began to analyze it in real-time to identify potential malicious activity.
After HIH informed them about this different application of their platform, OmniIndex decided to optimize their native AI specifically for this purpose so that it could work ‘out the box’ for new users. As a Small Language Model, this was simple to achieve because Boudica does not require extensive training in order to do a new task, instead it just needed a specialist thesaurus added so it knew what to look for. The challenge, however, was knowing how the AI and ML would actually be used to gain the most valuable insights and so OmniIndex worked with HIH’s analysts, as well as other experts, to provide the required automation and features so the financial fraud detection could work with full automation and simplicity.
During this process, OmniIndex realised that there was additional intelligence within their web3 data store that could aid fraud detection. This is because as data is stored in an immutable blockchain, all logs are fully auditable in a way that cannot be manipulated; including details of transactions, attempted access, sign in times etc. This fully encrypted information was already being used within the platform to provide threat and compliance intelligence, and so OmniIndex also made it accessible as part of the same AI and ML fraud analysis to provide a comprehensive picture of activity.
OmniIndex now offers this AI fraud detection to all new customers as part of their PostgresBC Data Platform and it will also be included in their Web3 File Store Dropblock which is set to be released later this year.
