Pundi AI Transforms Datasets into Tradeable Crypto Assets

Singapore-based Web3 firm Pundi AI has introduced a novel feature called Data Pump, designed to convert AI training datasets into tradable digital assets on the blockchain. This new functionality allows contributors to tokenize their datasets as BEP-20 tokens, enabling them to earn royalties and trade on decentralized exchanges such as PancakeSwap.

The rollout of Data Pump comes at a time when the AI training data market is expanding rapidly, with forecasts estimating its value to reach approximately $9.58 billion by 2029. Despite this growth, data monetization remains relatively underdeveloped, with less than one-fifth of companies currently engaged in any initiatives to derive value from their datasets. Pundi AI has positioned Data Pump as a response to this disconnect between data’s rising economic importance and the limited opportunities available to data contributors under conventional AI development models.

According to platform details, the system has already facilitated the minting of over 4.5 trillion data tokens, spread across more than 122,000 datasets contributed by over 28,000 users. These datasets undergo tokenization through a process that verifies ownership using NFTs, creating a transparent and secure pathway for contributors to maintain control while participating in a broader data economy.

A key innovation behind Data Pump is its use of a bonding curve pricing model, which introduces market-driven valuations for tokenized datasets. This approach ensures that token prices respond to supply and demand dynamics rather than being arbitrarily set. The model also includes anti-front-running mechanisms to safeguard the integrity of trades, while allowing token holders to benefit from increasing dataset popularity through a liquidity-based pricing structure.

Through this model, Pundi AI aims to address one of the longstanding challenges in the data market: how to determine the fair value of a dataset before it has been utilized. By embedding pricing logic into smart contracts, the platform enables automated royalty distributions, thereby reducing operational friction and ensuring continued compensation for data creators as their assets are reused.

The tokenization of data assets also fits within a broader trend in Web3, where decentralized models of ownership and compensation are seen as a remedy to the centralized control that characterizes traditional tech infrastructure. In conventional systems, tech giants often extract immense value from AI development while offering minimal rewards to the data providers who contribute to training models. Pundi AI’s mechanism attempts to realign this imbalance by transforming passive datasets into dynamic, monetizable assets.

An additional challenge in the data monetization ecosystem has been data quality. A majority of organizations, nearly 56%, have cited poor data quality as a major hurdle to monetizing their datasets effectively. Pundi AI’s platform integrates AI agents to pre-inspect data quality before purchase. This adds a layer of transparency to the process, allowing developers to assess utility and accuracy upfront rather than discovering quality issues post-integration.

By enabling datasets to be independently verifiable and backed by on-chain reputation mechanisms, the platform introduces community-led validation into a traditionally opaque marketplace. This not only enhances buyer confidence but also embeds accountability through immutable records of provenance and permission history.

As part of Nvidia’s Inception program, Pundi AI has been actively expanding its offerings beyond Data Pump, including the development of an omnichain layer and a decentralized data marketplace. The company’s efforts represent a shift toward more inclusive and transparent data ecosystems, where contributors are equitably rewarded and AI development is aligned with ethical and sustainable principles.

Through Data Pump, Pundi AI is not only unlocking liquidity for previously illiquid digital assets but also laying the groundwork for a fairer data economy—one that acknowledges and compensates the true value creators in the AI ecosystem.