Cape Privacy raises $20M to enable data science operations on encrypted data

Enterprise

Join Transform 2021 this July 12-16. Register for the AI event of the year.


Cape Privacy, which is developing a privacy-preserving platform for collaborative data science, today announced that it closed a $20 million series A led by Evolution Equity Partners. CEO Ché Wijesinghe says that the proceeds will be used to support growth as Cape Privacy develops new technologies for secure machine learning.

AI promises to transform — and has transformed — entire industries, from civic planning and health care to cybersecurity. But privacy remains an unsolved challenge, particularly where compliance and regulation are concerned. Banks, health providers, and even retailers can run into problems when collaborating on AI and machine learning research involving sensitive or proprietary data, like patient records, financial documents, and supply chain details.

Cape was founded in 2018 by Gavin Uhma, the cofounder and CTO of GoInstant, which was acquired by Salesforce in 2012. Cape’s combination of privacy, machine learning, and cryptography enables encrypted data-sharing, helping teams in compliance, legal, and risk management work with each other and third-party vendors.

“Today many financial institutions access the same publicly available data from data providers like Nielsen and Bloomberg — but they all want a better edge,” a spokesperson told VentureBeat via email. “Non-public data sources such as those from retail and credit card companies would greatly enrich their models. Yet concerns around confidentiality on both sides have prevented this collaboration. Many data providers are interested in finding new channels to monetize their data, but they can rarely get it past their internal legal and compliance teams.”

Cape’s open source software integrates with data science and AI infrastructure to provide a workflow guiding contributors toward building custom projects and policies. Cape enables developers to decide on the placement of tools in relation to data storage and pipelines, ensuring data access, privacy, and monitoring meet each product’s requirements. Moreover, it allows stakeholders to set monitoring and auditing configurations so that all parties receive logs for review, approval, or amendment.

“Cape Privacy’s platform … ensures privacy by default. With Cape as the broker, data providers are only renting data instead of selling it. This is a significant point because companies that lose control of their data can get in trouble,” the spokesperson said. “Once a data model is enriched using encrypted data on the Cape cloud, the transaction between buyer and seller ends and the data is returned. Now the data subscriber can enrich its data for better business outcomes, and the data provider can securely monetize its data.”

Encrypted data analytics

Cape’s platform is underpinned by tf-encrypted, the company’s suite for experimenting with private machine learning on top of Google’s TensorFlow framework. Tf-encrypted enables training, validation, and prediction over encrypted data. The data remains encrypted during the workflow, meaning that AI models can be hosted in the cloud without decrypting the training data or outputs.

Seventeen-employee Cape, which claims to have two major clients and “half a dozen” in the pipeline, isn’t the first to advance a privacy-preserving data science approach. Companies including Enveil, Cosmian, Duality Technologies, and Intel are investigating homomorphic encryption, a form of cryptography that enables computation on file contents encrypted using an algorithm so that the generated encrypted result exactly matches the result of operations that would’ve been performed on unencrypted file. Using homomorphic encryption, a “cryptonet” can perform computation on data and return the encrypted result back to a client, which can then use the encryption key to decrypt the returned data and get the actual result.

Homomorphic encryption libraries don’t yet fully leverage modern hardware and are at least an order of magnitude slower than conventional methods. That said, newer projects like the accelerated encryption library cuHE claim speedups of 12 to 50 times on various encrypted tasks over previous implementations. And HE-Transformer, a backend for nGraph (Intel’s neural network compiler), delivers leading performance on some cryptonets.

New investors Tiger Global Management, Ridgeline Partners, and Downling Lane participated in Cape Privacy’s series A together with existing investors Boldstart Ventures, Version One Ventures, Haystack, Radical Ventures, and Jevon MacDonald. Additional investment came from Coinbase cofounder and board member Fred Ehrsam, the Tokyo Black Fund, and Sand Hill East. To date, New York-based Cape Privacy has raised over $25 million.

“We are excited at reaching this company milestone,” Wijesinghe told VentureBeat via email. “Cape’s technology will be a defacto standard for privacy preserving machine learning. Building on our success in the financial services industry, we have already had great interest from Health and Life Sciences companies for potential drug discovery and genomics research use cases. In addition, there is clear demand for this technology for collaboration on machine learning model development across government agencies for counter-terrorism programs.”

VentureBeat

VentureBeat’s mission is to be a digital town square for technical decision-makers to gain knowledge about transformative technology and transact.

Our site delivers essential information on data technologies and strategies to guide you as you lead your organizations. We invite you to become a member of our community, to access:

  • up-to-date information on the subjects of interest to you
  • our newsletters
  • gated thought-leader content and discounted access to our prized events, such as Transform 2021: Learn More
  • networking features, and more

Become a member

Products You May Like

Leave a Reply

Your email address will not be published. Required fields are marked *