Quick take:
- The fundraising also attracted participation from Borderless Capital, Anagram, Foresight Ventures and Karatage.
- Fraction AI is currently in a closed testnet and is scheduled to launch its public testnet in January.
- The company uses a hybrid model for data labelling, combining both human inputs and AI agents.
Fraction AI, a decentralised AI platform combining both human inputs and AI agents for data labelling has raised $6 million in a pre-seed funding round led by Spartan Group and Symbolic Capital.
The Fundraising also attracted participation from Borderless Capital, Anagram, Foresight Ventures and Karatage, with Polygon’s Sandeep Nailwall and Illia Polosukhin or NEAR Protocol joining as angel investors.
The fundraising was structured as a simple agreement for future equity (SAFE) with token warrants, Fraction AI founder and CEO Shashank Yadav told The Block.
The company is building a decentralised data platform that leverages both human input and AI agents to train AI models.
“Among AI’s three core elements — data, compute and models — data remains the most obscure and tightly controlled,” Yadav said, adding that his company is determined to change that and level out the playing field by empowering anyone to train high-quality AI models.
The platform features three main types of contributors, stakers — who earn yield by staking Ether or liquid staking tokens, builders — who provide human insights and detailed instructions in text form, and Judges — the specialised LLMs tasked with evaluating the agent outputs against pre-determined criteria.
Built on the Ethereum network, the platform has plans to launch on NEAR and several other Ethereum L2s, Yadav said.
Stay on top of things:
Subscribe to our newsletter using this link – we won’t spam!
