Capnion is all about making advanced technologies accessible, and towards this end we have created and continue to maintain the Ghost PII software development kit. Ghost PII is designed to make advanced privacy enhancing technologies (PETs) available to your everyday data scientist or software engineer working in Python without too much additional training or stress. This GitHub repo is a great place to go for more detailed information or you can get the high-level story below.
If you feel burdened by the sensitivity of the data behind your application development or data analytics project then we might be able to make your life easier.
Data privacy shouldn’t be a burden to the people working in the back of the office, it should be an asset to the people working in the front of the office. This transformation is what Capnion and Ghost PII can provide. Below you can learn…
Who uses Ghost PII? What do they use it for?
Why should I use Ghost PII? How might Ghost PII improve my $$$ bottom line $$$?
Why Ghost PII and not something else? What does Ghost PII do uniquely well?
Having just fallen in love, how can I get started using Ghost PII?
What is Ghost PII? Who uses it and to do what exactly?
At Capnion our mission is to make a higher data privacy standard easily accessible to software engineers and data scientists working in the open source Python ecosystem. Our design goal is always that someone who is already working with Python can pick up our tool Ghost PII with minimal retraining, or even with existing code following small changes to how data is ingested. If you are working with Pandas or Django, for example, you will find Ghost PII to be an agreeable and empowering addition to your stack.
Ghost PII makes it possible, via cutting-edge and qualitatively new encryption techniques, to build applications and analytics pipelines that function on securely encrypted data all the way through with no decryption at any point – this is the ultimate pervasive encryption technology. For an application, we sometimes jest that this amounts to “even end-to-endier encryption” in that not only the primary data but also the metadata remains encrypted at all times with no plaintext data on the application server at all. For an analytics pipeline, this means giving your team the ability to see things they need to see like summary statistics, regression coefficients, machine learning predictions and performance, etc. without ever giving them access to the underlying raw data.
Why should I use Ghost PII? What benefits will my business enjoy if we start using it?
Pervasive encryption is of course great for information security, but it may be in compliance that you find Ghost PII most transformative. This is could be satisfying requirements for avoiding personal information appearing in plaintext under GDPR, FERPA, or HIPAA. It could also be obeying your internal rules on sensitive data in test and development environments, or it could simply be satisfying the concerns of a data partner and putting their mind at ease.
This may unlock business opportunities that would be inaccessible otherwise. For example, at time of writing data transfer out of the European Union is in turmoil following the Schrems ruling, but recent guidance states transfers are permitted for purposes that do not require data in the clear. Vastly expanding the space of tasks that can be performed without decrypting is exactly what Ghost PII offers. If your business requires sensitive data from customers, this could mean Ghost PII allows you to access the European market when you would otherwise be excluded. There is a quick demo following this storyline in the video below.
Just think about high-level GDPR requirements like privacy-by-design… What could satisfy this requirement better than complete elimination of plaintext personal information throughout a pipeline?
You know the value in your data. We bring data privacy into a new level of harmony with your other goals and let you go get it.
I am drowning in vendors. What can Ghost PII do that I won’t find anywhere else?
Ghost PII doesn’t require you send any data to Capnion – we send information to you via our key hosting service that our Python client interprets to achieve the feats described above. You are not going to pick up any extra breach risk from us, and in general Ghost PII is an and-plus technology that you can use in addition to other security you may have already.
Especially unique, even compared to other cutting-edge encryption products, is the rigorous audit trail Ghost PII produces describing how and when your data is used to a cell-by-cell level of detail. It produces this audit trail accurately and exhaustively even when data leaves your network and your direct visibility. The integrity of this audit trail is enforced by cryptography – there is no way to extract value or information from the encrypted data without generating it.
Once again, the regulatory value of this audit trail is considerable. Tracking such information is a frequent explicit requirement, and heaven forbid anything of any sort go wrong it will only help you soothe regulators if you can show them you went an extra mile keeping tabs on your data.
Ghost PII can be used for much than the personally identifiable information (PII) in its name. Ghost PII can be used, for example, to perform machine learning on encrypted data so that nothing needs to appear in plaintext but the insights that drive your business. However, at Capnion we feel improving privacy around the handling record linkage on personal information is really where the important action can be found, and Ghost PII is best-in-class for performance, security, and flexibility when applied to text data like names, phone numbers, and addresses that constitutes your average embarrassing data breach. Ghost PII can be used to perform fuzzy matching, data cleaning, and similar operations on encrypted data directly that would require access to plaintext under tokenization or cryptographic hash -driven approaches.
At a high-level Ghost PII is particularly well-differentiated from products that have come before when there are data quality concerns or in situations like cross-firm data pipelines that make it harder for one central authority to exercise control of data governance.
I love it. How can we get started?
If you read something here that appealed to you, !email acmueller@capnion.com right now!. We know this a very novel technology and we are ready to go the distance educating you – we can provide you with case studies, videos, access to demo applications and live sandboxes, and of course answers to your questions. Today should be the day that you help your organization make data privacy the empowering asset it should be.