Fraud detection/Anti Money Laundering - Using CipherMode, external organizations can share their data without disclosing anything about the individual data elements. An AI model can now be trained globally across multiple datasets to pick up more fraud than can be found using a local model. In our pilot, CipherMode enabled a global model that was able to increase detection by 10-25% over the local model.
Customer Data Sharing - Organizations would love direct customer spending and sentiment data so that they can more specifically tailor financial products. Using CipherMode, a bank can enable customer trust since the data is always encrypted while the bank is using it for analysis.
Insurance Data Sharing - Insurance companies typically share as little data as possible since their data is their competitive advantage. However, the more data they have, the better their actuarial models can perform so there is incentive to share data more widely. CipherMode can be that trust layer to ensure insurers, re-insurers, brokers and risk modelers can share data more widely to offer better priced products all while protecting their crown jewels and ensuring customer privacy.
Providers 3rd-party data sharing - Providers are goldmine of healthcare data from patient to wait times to incidence of heart attacks in various populations - providers have it all. Many companies are interested in this data to train AI models or run inferences to extract insights but due to regulatory requirements and the conservative nature of providers, this data is often not easily shared. CipherMode can facilitate this sharing by encrypting the data when a 3rd party wants access and only decrypting once the data is done being used.
Drug discovery and clinical trials - Drug discovery requires many parties participating in order to bring new products to market quickly. CipherMode can be a trust layer that enables multiple parties to share data for analysis while maintaining confidentiality and ensuring PHI is kept private.
Predictive maintenance - Infrastructure technology providers increasingly want to predict when portions of their infrastructure may go offline and impact their customers. In order to build models that can predict outages, they need their customers’ data so that their models have enough events to start picking up patterns. CipherMode can play the role of facilitator and connective tissue to build trust between providers and their customers so that the end user feels the least impact on their service
Multi-jurisdictional data sharing - Tech companies are growing into new geographies with various regulations that limit how data can be used. Organizations today are blocked from training a model on sensitive data in multiple jurisdictions because the only way data can leave a jurisdiction is for it to be encrypted. CipherMode can be the facilitating tool to encrypt the data while enabling the AI model to train across global datasets.