🎉 Congratulations, you’re in.

You have been selected to participate in this Hackathon in partnership with Entrepreneur First, Hugging Face and Zama.

This page serves as your official welcome pack - here you can find everything you need to know ahead of September 26th.

<aside> 💡 Practical information

📍 Location: 8 Rue du Sentier, 75002 Paris

🥬 Any food restriction? Let us know here!

👾 Discord: launch on Monday September 23rd

🤝 Link to spreadsheet for teaming up

Night coworking recommendations

FAQ

</aside>

Privacy Preserving AI Hackathon - Slides.pdf

📣 Share your participation to this Hackathon

This Hackathon is highly selective, and you are among the 50 lucky ones who have secured a spot out of the hundreds of applications we received. This can increase your visibility on LinkedIn and make you more attractive to employers if you're seeking an internship or job. It could enhance the visibility of the event overall and consequently yours as well if you win this hackathon (which should be your goal!)

• Ideas of post
• Visuals to add to your post

📆 Hackathon Schedule

<aside> 💡

Thursday September 26th from 6:30pm - Intro to the hackathon, meet all participants and form your team at the Zama office with food and drinks.

</aside>

Friday, September 27th

Saturday, September 28th

🔍 Technical Info

😎 Zama

<aside> 👉

Get familiar with Zama’s open source Concrete ML library

What is Concrete ML?

Concrete ML is a Privacy-Preserving Machine Learning (PPML) open-source set of tools built on top of Concrete by Zama.

It simplifies the use of fully homomorphic encryption (FHE) for data scientists so that they can automatically turn machine learning models into their homomorphic equivalents, and use them without knowledge of cryptography.

Concrete ML is designed with ease of use in mind. Data scientists can use models with APIs that are close to the frameworks they already know well, while additional options to those models allow them to run inference or training on encrypted data with FHE. The Concrete ML model classes are similar to those in scikit-learn and it is also possible to convert PyTorch models to FHE.

Main features

• Built-in models: Ready-to-use FHE-friendly models with a user interface that is equivalent to their the scikit-learn and XGBoost counterparts
• Customs models: Concrete ML supports deep learning models defined by users, using either post-training quantization or quantization-aware training. These are developed by the user using PyTorch or Keras / TensorFlow.

Learn more about Concrete ML features in the documentation.

</aside>

🤗 Hugging Face

<aside> 👉

Some Hugging Face spaces already built with Concrete ML

🕵️ Encrypted anonymization:

https://huggingface.co/spaces/zama-fhe/encrypted-anonymization

🏥 Encrypted health prediction:

https://huggingface.co/spaces/zama-fhe/encrypted_health_prediction

💳 Encrypted credit card approval prediction:

https://huggingface.co/spaces/zama-fhe/encrypted_credit_scoring

🖼️ Encrypted image filtering:

https://huggingface.co/spaces/zama-fhe/encrypted_image_filtering

🙂 Encrypted sentiment analysis:

https://huggingface.co/spaces/zama-fhe/encrypted_sentiment_analysis

</aside>

<aside> 👉

Some Hugging Face endpoints already built with Concrete ML are:

• Synthetic dataset classification with a Logistic Regression

https://huggingface.co/zama-fhe/concrete-ml-encrypted-logreg

• Spam detection with a Decision Tree

https://huggingface.co/zama-fhe/concrete-ml-encrypted-decisiontree

• Iris classification with a QNN

https://huggingface.co/zama-fhe/concrete-ml-encrypted-qnn

• MNIST digit classification with a CNN

https://huggingface.co/zama-fhe/concrete-ml-encrypted-deeplearning

</aside>

You’ll use a dataset of your choice and build privacy-preserving Spaces/Endpoints running on upgraded CPUs, with free credits.

Please complete these steps to be able to use the free credits:

1. Send a request to join the event Hugging Face organization here.
2. Each team will create either one Space or one Endpoint as follows:

🗣️ Pitches