Part 1 - Proposal Submission
Name of Project: FELToken
Proposal in one sentence:
Creating a decentralized and more secure solution for federated learning while anonymizing data providers.
Description of the project and what problem is it solving:
FELToken is connecting owners of the data with scientists to train their machine learning models while preserving the privacy of the data. Current federated learning solutions rely on a central server. We are building a product with the simplest user experience possible while being decentralized. This allows anonymous participation of data providers and prevents malicious activities. Data providers get rewards for sharing their data and resulting models can be further sold. The exchange of all models is encrypted so only authorized parties can use the model. We would like to further improve our tool by connecting with Ocean infrastructure.
Grant Deliverables: (Target deliverables for the funding provided.)
- Having easy to use end-to-end demo running on testnet (MVP)
- Create demo script: listing data on Ocean + using compute-to-data
- based on this we will then create the full integration
- Write blog post about our experience and progress with Ocean
- Finding UX/UI designer
Which category best describes your project?
- Build / improve applications or integrations to Ocean
Are you applying for an Earmark?
- General
What is the final product?
The final product of FELToken consists of three main components: a set of smart contracts governing the federated learning process, a web application for simple interaction with smart contracts, and a tool for running a data provider node. With Ocean integration, users will be able to use the compute-to-data as a node. We would also like to add the possibility of further selling trained models through Ocean.
During development, we focus mainly on making the process of setting up the federated learning project as simple as possible. So that data scientists and interested parties can train their models without any extra knowledge about smart contracts and blockchain. Once the core components are working there are many possible ways for further extension. The tool can also act as a platform for further federated learning research.
Value add
- Usage of Ocean
- Viability
We already have a working demo of the core technology (smart contract and code for data provider). We also finished core parts of the web application: starting a new project, displaying the project dashboard and creating a new training model. Right now we are working on simplifying the connection of data providers. Once finished with that, we should have MVP which is worth sharing with others. After that, we will continue working on more improvements. We are especially interested in using Ocean’s compute-to-data solution which could replace our data provider code. This would allow seamless integration with Ocean marketplace.
We believe that this tool can drive the attention of more people working on machine learning and data privacy. We are also focusing on companies that will use this product with their real data. We believe that this could bring more people to the community overall.
Funding Requested: 10000$
Proposal Wallet Address: 0x77edDB82E5e9901aA494825bC362fA93120B892c
Have you previously received an OceanDAO Grant?
Y
Team Website: https://feltoken.ai
Twitter Handle: @FELToken
Discord Handle: Breta#9929
Project lead full name: Břetislav Hájek
Project lead email: info@bretahajek.com
Country of Residence: Czech Republic
Part 2 - Team
2.1 Core Team
Břetislav Hájek
- Role: developer
- Relevant Credentials:
- Background/Experience:
- AI developer: https://inventives.ai/
- PhD student at: https://www.nus.edu.sg/
Filip Masár
- Role: developer
- Relevant Credentials:
- Background/Experience:
- Kernel KB4 fellow https://kernel.community/
- Co-founder at http://avtr.gg/
- Front-end developer at https://avast.com/
Martin Ondejka
- Role: developer
- Relevant Credentials:
- GitHub: mondejka (Martin Ondejka) · GitHub
- LinkedIn: https://www.linkedin.com/in/martin-ondejka/
- Background/Experience:
- Python developer at https://www.kiwi.com/us/
- Software Engineer at https://vacuumlabs.com/
Part 3 - Proposal Details
Details of the proposal:
We are building software for federated learning projects. As such we need to provide an easy UI for companies/interested parties to set up their projects. Right now we are finishing the code and simplifying the user experience. After that we can publish the official MVP. Meanwhile, we want to start working on Ocean integration.
App will be live at:
Is the software open-source?
Y
If open-source, please specify the license. If no, please specify why not open-source.
GPL-3.0 License
Project software can be found at:
Are there any mockups or designs to date?
Y: There is this video used for hackathon submission explaining the project: Federated Learning Token - FELT - YouTube
Please given an overview of the technology stack.
Smart contracts are written in solidity with the use of OpenZeppelin contracts. There are three main smart contracts: token contract, project manager and project contract. Project contract is the main contract managing federated learning processes. We are planning to deploy them on the Polygon (MATIC) network.
The web application is written in TypeScript using React as a frontend framework. For communication with smart contracts, we use the Ethers.js library.
The data provider client is written in Python. Right now we support scikit-learn models. We will extend this for TensorFlow and PyTorch in the near future.
3.9 Project Deliverables - Roadmap
Any prior work completed thus far? Details?
Y: We started the project as part of Chainlink Fall Hackathon 2021 where we created the core contract and data provider code. Last month we received the OcenDAO grant and we worked on providing main functions (communication with smart contracts) through the web application.
What is the project roadmap?
- Simplify the client code
- Add a few extra functionalities to the web application for interacting with smart contracts.
- Start working on a demo using Ocean’s compute-to-data (this demo is mainly for us to understand how to use Ocean and plan the integration accordingly).
- Release MVP
- Add integration with Ocean (training using compute-to-data)
- …
What is the team’s future plans and intentions?
There are many possible extensions. As mentioned before we want to integrate with different tools and marketplaces out there. Once the core of the project is finished we want to focus on getting people to use the project (working as a support and fixing bugs as needed). We would like to start working with some companies to better understand their needs related to federated learning. We would also like to attract researchers to develop new algorithms for federated learning on blockchain. Run machine learning competitions in federated learning settings.
3.10 Additional Information
Any additional information
- We were one of the winners of “On the Rise” category in Chainlink Fall Hackathon 2021
- Receive grant as new project during OceanDAO round 13
Screenshots from current application (goal of Round 13)
Blog post from end of Round 13