Aether Holdings Partners With OORT To Create Human-Verified Datasets For Fintech AI

Fintech research and media platform Aether Holdings has teamed up with the decentralized data infrastructure startup OORT to create a new entity to address the chronic shortage of high-quality, verifiable training data for financial AI models.

TechStartups has learned that the new entity is called Aether DataHub, and it plans to leverage Aether’s extensive user base and OORT’s technology to create a decentralized platform that enables communities to gather, label, and verify financial datasets for training targeted fintech AI models.

The AI industry has made massive leaps in terms of computing power and model architecture, but even with these developments, the capabilities of AI models are still largely determined by the quality of the data they’re trained on. In some industries, data isn’t a problem. There are billions of free images posted online, which is why training image generators is a fairly simple business these days. But in the case of financial services, getting a hold of accurate data is harder than it looks.

Data is everywhere, but trust is scarce

Aether says there’s a drastic shortage of high-quality, domain-specific financial data online. While that may come as a surprise, the problem is that most financial data online lacks the “institutional-grade validation” needed to train specialized AI models. Financial services firms such as banks and hedge funds may have millions of dollars at stake, and so they cannot tolerate the risk of AI hallucinations, no matter how small that risk is. What’s needed is a way to verify financial data, so they can trust the models it’s used to train.

This is the challenge Aether DataHub is setting out to conquer, and it’s doing so by combining its global financial media footprint with OORT’s decentralized data infrastructure.

The idea is to use OORT’s existing DataHub platform, which enables communities to collect, process, label, organize, and monetize high-quality datasets for training AI models. By offering contributors crypto rewards, it incentivizes its users to handle the manual grunt work and ensure that the datasets are prepared and verified for us. Aether DataHub has the powerful advantage of being able to tap into Aether’s 400,000-strong community of global investors. These users, many of whom possess considerable financial expertise, will serve as a “distributed intelligence layer” and provide the institutional guarantees financial firms need.

Aether has put its name to the new entity, and it says it will maintain majority governance and focus on driving adoption, while OORT will handle the technology that powers it. According to Aether CEO Nicolas Lin, it gives the company a chance to “own and operate a critical layer of financial AI infrastructure.”

Accelerating fintech automation

For years, AI developers focused on building bigger, better models with more parameters to outmuscle the competition, but nowadays there’s a growing recognition that sheer power alone is not enough.

What’s needed is more targeted data that can be used to train AI models that are especially good at specific tasks, and that’s where Aether sees an opportunity. By decentralizing the workforce needed to supply and prepare that data, Aether not only reduces costs but also builds a moat in the form of rich, human-validated data that cannot be matched by competitors who simply scrape the web and hope for the best.

Aether is evolving into more than just a media company and a fintech firm. It’s positioning itself as a key infrastructure provider, just in time for the “agentic AI” gold rush. While automation has been accelerating elsewhere, financial services firms have been held back by regulatory concerns, so Aether’s verifiable datasets should be highly appealing.

Aether’s biggest challenge will be governing its distributed network at scale, but its community of financial experts should have sufficient expertise to ensure it maintains the highest institutional standards.