MOSTLY AI Raises $25M to Bring Synthetic Data to Everyone’s Life

MOSTLY AI Raises $25M to Bring Synthetic Data to Everyone's Life

Watch the live videos from the Low-Code/No-Code Summit to learn how to innovate and improve efficiency by scaling and upskilling citizen developers. Check it out now.

Austria-based MOSTLY AI is a startup that simulates synthetic data to assist with AI model testing and training. It has today announced that it had raised $25 million from a series B round of funding from Molten Ventures. The company will use this funding to expand its efforts in laying the stage for unbiased and responsible AI, bringing on new talent, and expanding its presence throughout Europe and North America.

In any modern enterprise, the most significant challenge in leveraging data for AI/ML is to ensure the privacy of their customers — the sources of data in the first place — as well as remove the possibility of bias based on social or historical inequities that exist on the data. Companies often have difficulty when it comes to tackling these two issues and find themselves in the position of having to pay penalties for privacy violations (under laws like GDPR) or create an algorithm that is unjust concerning one or more parameters.

Also Read: Celestial AI lands $56M to create lights-based AI accelerator chips

MOSTLY AI opens up its synthetic data platform

To address the issues, data scientists Michael Platzer, Klaudius Kalcher, and Roland Boubela started Mostly AI in 2017. The company uses AI to build a real and representative synthetic dataset that preserves the information necessary to complete the data value chain, from AI model training to advanced analysis to software testing and coding – but without individual data points.

This provides data engineers and scientists with as accurate but completely anonymous data for them to use.

The solution is based on an advanced deep neural network that generates generative patterns with an integrated privacy mechanism. It recognizes important patterns, patterns, and variants from original data and then recreates these patterns using a group of fictional characters. This gives an imitation of the secure, de-biased original, and as valuable as the original dataset, that can reflect patterns and behavior with as high as 99percent accuracy.

Also Read: What Are Some Ways You Can Promote a TikTok Account?

Additional benefits

Alongside ensuring that AI/ML projects are fair and secure in terms of privacy, the platform also increases the time it takes for enterprises to get to data. This is because synthetic data can be produced rapidly and in abundance, unlike the original datasets. The company says its technology has already been shown to decrease time-to-data by 90 percent, saving the largest companies around $10m per year on data provisioning and internal overhead and increasing data availability by up to 85 percent.

With the most recent funding round, which included an involvement of Earlybird, 42CAP, and Citi Ventures, Mostly AI plans to expand its presence across Europe and the US and Europe and expand its customer base within the insurance and banking sector. It is reported that the company has already signed numerous Fortune 100 banks and insurers.

But it’s not the only one in this field. Tonic.ai, Synthesis AI, Hazy, and Gretel are several other companies working to develop fake datasets that will help businesses accelerate the speed of their AI projects.

Also Read: CommerceIQ Raises $115M in Series D Funding

Demand for artificial intelligence AI

Due to the rising concerns and regulations surrounding data privacy and the growing need for data-driven solutions, synthetic data is predicted to be a major driving force for companies within the next few years. Based on Gartner estimates that in 2024 60% of the data utilized in the development of AI and analytics applications will be created synthetically. In a different study, 9 out of 10 technical decision makers who use vision data claimed that synthetic data is a brand new and exciting technology essential to stay ahead of the trend.

“2022 is expected to be an era of artificial data. Synthetic data can help solve many of the most problematic problems when it comes to AI. It removes the concerns regarding data security, and it can be freely designed and created to boost AI projects and allows companies to expand and remove bias from their data sets.” Tobias Hann, CEO of Mostly AI, said in an announcement. “We’re very excited about the possibilities of artificial data, and we are pleased we are thrilled to partner with Molten Ventures, which shares our goal of fundamentally altering how companies interact about data.”

Chris Evan was born in Dubai and raised in Montreal. He studied Computer Science and was so pleased with computer languages. He began writing after obsessing over technology.

Leave a Reply

Your email address will not be published. Required fields are marked *