Allows Multiple Parties with Distributed Data and Models to Collaborate Safely
Habu, the Global Innovator in Data Clean Room Software, today announced the release of its CleanML product, enabling two or more partners to bring distributed models and data together under a protected and governed environment that allows both parties to benefit. A natural evolution for today’s data clean room technology, CleanML enables data science teams to transform the utilization process by identifying other valuable data or model sets to determine new collaboration opportunities.
Marketing Technology News: MarTech Interview with Cailin Pitcher, VP of Marketing at Paubox
“Imagine a scenario where one partner has a machine learning model and another party has data with a mutual opportunity to connect the two for inference”
CleanML facilitates a place for innovation and collaboration for data science teams and their partners alike. It can be utilized across a variety of business and industry verticals including retail, consumer packaged goods (CPG), auto, travel, financial service, healthcare, and more.
With CleanML, multiple parties can provide distributed raw materials — such as a model, model-training code, or dataset — for machine learning, with each respective partners’ assets remaining safe in the protection of its own clean room. Habu also provides the Habu CleanCompute module, a secure run-time environment where private datasets are joined for analytics or modeling. Privacy-preserving techniques prevent each party from seeing the other’s model or data, while both benefit from the output and results.
Marketing Technology News: NielsenIQ and Qualtrics Announce a New Partnership to Help Brands Drive Sustainable Growth
“Imagine a scenario where one partner has a machine learning model and another party has data with a mutual opportunity to connect the two for inference,” said Matt Karasick, CPO at Habu. “The company with the data can’t share the data, and the company with the model can’t share their model. But both organizations can benefit from the output of the collaboration. This is exactly why Habu developed the CleanML and CleanCompute intelligence modules whereby distributed data, machine learning models, and/or containerized microservices can connect in safe and secure ways.”
For users who are not data scientists, CleanBI is the industry’s first low code, no code solution for distributed queries while preserving privacy and minimal movement of data. The collaboration process typically begins by defining the questions that the various partners want to answer about what is happening in a given situation or what is possible in a situation. CleanBI often begins with use cases for one particular department, such as marketing. As data science teams get more involved in the process, the Habu software solutions are applied horizontally across many departments and business functions.
Marketing Technology News: Blockchain marketing Agency Hype Partners Acquires DeFi Consultancy Agency0x