c-Hive

c-hive logo

Inspiration

Privacy is one of the most challenging aspects of our current data-driven world. Data-driven individuals and small businesses can be some of the most vulnerable groups when it comes to handling private and sensitive data, since they tend to rely on cheap and available means of storing and sharing data. c-Hive is designed to help vulnerable groups handle data without leaks and with full customer data anonymity! c-Hive can be used to ramp up the computing power for a wide range of applications using healthcare, customer, or financial data where personal data needs to be kept private.

What it does

Using a technique called Homomorphic Encryption, it is possible to encrypt data, perform calculations on the encrypted data, and then decrypt afterwards to get the true (unencrypted) result. This allows for the outsourcing of the calculation step, without worrying about data leakage. cHive links 2 participating nodes (data owner and data receiver) and allows the less computationally powerful data owner to send its sensitive data to a powerful receiver. Computations are securely carried out before sending the results back to the data owner, who is the only one able to decrypt.

How we built it

We used tenSEAL, a python package built on top of Microsoft SEAL, to handle the encrypted operations. We handle the P2P aspect using basically only the python 'sockets' and 'p2p' packages

Challenges we ran into

Understanding how to choose secure parameters for the encryption was tricky, as well as just being able to ping between 2 computers, since we had never written any networking code before

Accomplishments that we're proud of

Building this proof of concept!

What we learned

We learned more about encryption techniques, networking code and sockets, and that there is no need for small businesses and individuals to pay large cloud computing fees to ensure the privacy of their data analytics!