Skip to content

RnDVerse/AI4DeSci

Repository files navigation

description
[This project is initiated and submitted for Akashathon2 17th April- 17th May 2024]

AI4DeSci

AI4 DeSci Logo

Vision: Sustainable Infrastructure for AI-Empowered Open & Collaborative Decentralized Science

Mission:

[1] Build Infrastructure for Accelerating DeSci

[2] Develop AI4DeSci Use-Cases with Scientific Impact

[3] Promote DeSci Through Innovative Model and Actions

Build Exclusively on Akash.network

Akash.network

---------------------------------------------------------------------------------------------------------------------

AI4DeSci Ecosystem composed of 2 Core Platforms :

[A] DeSci Collaborative Cloud

[B] Science Crowdsourcing

AI4DeSci Ecosystem Flywheel

Architecture

Current Architecture of AI4DeSci

Links:

Website: ai4desci.com

Infra Overview: ai4desci.cloud

Documentation : docs.ai4desci.cloud

Collaborative Cloud : hub.ai4desci.com

Science Crowdsourcing: app.ai4desci.com

1st Use case : app.plantamusica.cloud

To give a quick idea what its going to look like, currently it is composed of mainly two parts. First part is for DeSci Cloud Collaboration. Imagine Google Colab for Decentralized Science on Decentralized Cloud. Differing from the traditional Colab platform, here collaboration could be Human-Human, Human-AI-Agents, also between AI-Agents. Second Part is for Science Crowdsourcing to generate high-quality labelled data for AI, think about roboflow.com for annotation and labelling of DeSci & open-science related project data. Due to time restriction, most effort for use case and demonstration will be build for the above two main part. Marketplace will be a natural third component that will slowly follow through after the Hackathon. Template, tools, and auxiliary services will be what offered and traded/rented through the marketplace.

About

Building Infrastructure for AI-Empowered Open & Collaborative Decentralized Science

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published