- In a pre-seed fundraising round headed by Spartan and Symbolic, Fraction AI has raised $6 million.
- The goal of the crypto-AI startup is to decentralize data labeling.
Spartan Group and Symbolic Capital co-led a $6 million pre-seed investment round for Fraction AI, a crypto-AI firm that aims to decentralize data labeling.
Fraction AI announced on Wednesday that Borderless Capital, Anagram, Foresight Ventures, and Karatage are among the other investors in the round. Angel investors that are “close advisors” to the business, such as Illia Polosukhin of NEAR Protocol and Sandeep Nailwal of Polygon, also participated in the round.
According to Yadav, Fraction AI finished the pre-seed round in September after starting to raise money in April of this year. The round was set up with token warrants and a simple agreement for future equity (SAFE).
Fraction AI: What is it?
Fraction AI is a crypto-AI firm that was founded in February with the goal of decentralizing data labeling. In order to train AI algorithms to identify patterns and produce precise predictions, data labeling entails assigning meaningful labels to unprocessed data, such as text, audio, or images.
Data is still the most elusive and strictly regulated of AI’s three fundamental components: computation, models, and data. In order to level the playing field and enable anyone to train superior AI models, we set out to fix that.
Fraction AI labels data using a hybrid methodology that combines AI agents and human insights. The three primary users of the platform will be judges, builders, and stakers.
By staking ether liquid staking tokens (LSTs), such as Lido staked ether (stETH), stakers will receive rewards. Entry fees paid by builders will be the source of their income, with stakers receiving 5% of each entry fee.
By offering human insights or comprehensive instructions in text format—without the need for coding—builders will generate agents. They will use LSTs or ETH to finance their agents so they may participate in challenges. In order to participate and produce the best data possible, builders will be required to pay a nominal admission fee. The entrance fee pool will be used to reward the top three agents out of five in each competition. The payments will be enhanced by a performance-based multiplier based on the results of specialized large language models (LLMs).
The stakers’ pool provides the additional multiplier-based payment, which guarantees that underperforming agents finance stakers while geometrically increasing payouts for top-performing agents.
Specialized LLMs known as judges will assess agent outputs in relation to predetermined standards. Judges must stake Fraction AI’s native FRAC tokens in order to take part.
Timeline for the Fraction AI mainnet and token launch
According to Yadav, Fraction AI is mostly based on Ethereum and is presently operational on a closed testnet with more than 60,000 users. The mainnet is scheduled for release by the end of the first quarter or early second quarter of 2025, while the public testnet is anticipated to begin next month.
According to Yadav, the FRAC token will also be introduced nearer the mainnet. According to him, the token will be used to safeguard a network of judges who will employ slashing and staking mechanisms to assess agent outputs, guaranteeing a fair and high-quality review.
Fraction AI intends to debut on NEAR in addition to several Ethereum Layer 2 networks, even though it is presently mostly created on Ethereum.
Currently, the San Francisco-based project employs eight people. Yadav intends to maintain the team’s size in the near future.
VC capital is still being attracted to crypto-AI businesses. Nonetheless, investors continue to express skepticism, cautioning that many of these firms may fail because they follow trends without having distinct value propositions.
Disclaimer : This article was created for informational purposes only and should not be taken as investment advice. An asset’s past performance does not predict its future returns. Before making an investment, please conduct your own research, as digital assets like cryptocurrencies are highly risky and volatile financial instruments.
