The narrative of WorkML.ai began with the partnership of Michael Bogachev and Denis Davydov in 2020 at a thriving Ukrainian startup, which was later acquired by the largest logistics company in the UAE. The unfolding conflict in Ukraine in 2022 forced them to relocate, a move that dramatically shifted their career trajectories and personal lives. In 2023, as a result of their travels across Europe, their paths converged again in Budapest. It was there, amidst the backdrop of their enforced displacement, that they conceptualized the core idea behind WorkML.ai.

As they explored potential ventures, Michael and Denis drew heavily on the prevailing trends in AI and cryptocurrencies. Denis brought a wealth of experience from his tenure with American cryptocurrency firms between 2022 and 2023, as well as his involvement in AI and crypto startups from 2016 to 2019. Meanwhile, Michael had been applying AI to enhance logistics systems since 2016. This blend of expertise and their shared experiences of adaptation and resilience formed the foundation of WorkML.ai, setting the stage for a project that would leverage cutting-edge technology to create impactful solutions in the AI space.

https://www.youtube.com/watch?v=Pmkcp8Mk0bg

Drawing on their expertise, they pinpointed key challenges in developing large AI models. 

The first major hurdle involved managing large datasets, an issue effectively addressed by Nvidia, which saw its stock prices more than double in 2023 following the launch of their new accelerators.

The second challenge, less apparent and only recognizable by those deeply involved in AI model training, is the meticulous preparation of Metadata. This metadata is crucial because it goes hand-in-hand with the main data fed into the models.

What is Metadata?

Metadata is a key element that allows the neural network to make an interpretation of what is visualized, voiced, or written, and how it relates to other objects.

You can learn more about this information in the WorkML.ai project Whitepaper.

Preparing Metadata is recognized as a complex task

To build a new neural network, it must be trained from the ground up with a vast quantity of data; using pretrained networks isn't an option as fresh training is required for each instance. Developers need both the data and the Metadata that describes this data. The accuracy of the Metadata and the volume of data utilized in training directly influence the intelligence and precision of the neural network's predictions.

For practical training of neural networks in tasks like animal image recognition or image generation, it's necessary to input tens of millions of images. Each image must be paired with Metadata that indicates where each animal appears on the image, marked using shapes such as rectangles, polygons, fills, or skeletons.

The complexity involved in the annotation process

The annotation process is intricate, particularly evident when considering that for 10 million images, between 30 and 40 million Metadata units are required. This is because each image may contain between one and ten or more objects, each needing identification. The precision of object marking, such as using polygons for more accuracy compared to rectangles, significantly impacts the effectiveness of the trained neural network.

It's clear that the demand for Metadata often surpasses the data itself. While data can be easily sourced in its original form, creating the necessary Metadata involves careful and methodical work.

For instance, an individual working consistently for 4.5 hours can annotate one item every two minutes, totaling 135 high-quality Metadata units daily. Over 21 workdays in a month, this results in 2,835 units.

To generate 35 million units of Metadata, one person would need about 12,345 months, or 1,028 years. A team of 100 would take just over 10 years to complete this, while 1,000 annotators could finish in a year.

The financial implications are substantial. The average cost for maintaining an office setup for one annotator is around $1,800 monthly. With 100 annotators, costs would escalate to $180,000 monthly over a decade, or $1.8 million monthly if employing 1,000 annotators, cumulatively reaching about $21.6 million for annotating 10 million images with 35 million Metadata units. See detailed research here.

Clearly, Metadata creation is a demanding process, requiring significant time and financial resources.

The founders at WorkML have developed a solution to this issue!

The solution involves creating an employment hub on the WorkML platform. Here, people worldwide can complete onboarding courses to join the workforce as annotators (annotator use-case) and data validators. This strategy aims to recruit tens or even hundreds of thousands of annotators for various tasks. Furthermore, companies can set up their own annotation departments via the WorkML platform, integrating outsourced annotators into their teams. This approach is expected to significantly improve the quality and efficiency of annotations while reducing costs by about tenfold.

This innovation is as vital to the AI industry as Nvidia's accelerators are.

Additionally, toreduce costs and feed, the project incorporates cryptocurrencies for transactions. Crucially, it will launch its own token, WML, which will be utilized for internal payments and compensating annotators.

One of the interesting features of the project is the crowdfunding approach of raising initial funds using cryptocurrencies for real-world projects with real business value. This approach allows you to create additional value for the crypto community and at the same time attract investment in the project.

To apply this method, it was necessary to think through a system of rewards and values for the crypto community that would motivate investors.

Such mechanisms include staking (something similar to a bank deposit), mining (carrying out physical work) and airdrops (tokens distribution events). To meet the needs of staking, we pledged 8% of the total amount when mint the token, which is enough for more than 10 years of payments. Mining in our case implies the real work of the annotators.

The token features:

• Proof of Stake (PoS): Provides variable payouts, rewarding token holders based on the amount of their stake, incentivizing long-term holding and investment.

• Human's Proof of Work (H-PoW): Rewards annotators based on the quality and quantity of their work, directly influencing their compensation. This mechanism aligns the incentives of annotators with the quality of data annotation.

• Human's Proof of Stake (H-PoS): An innovative feature that offers double payouts for those who reinvest their earnings obtained through Human's Proof of Work (H-PoW), significantly increasing rewards for active participants.

• Perpetual Discounts: Payments made with WML tokens for services on the platform receive perpetual discounts, enhancing the token's liquidity and appealing to users to transact using WML.

• Three-Level Referral System: A multi-tiered referral program rewards users who help expand the community by inviting new annotators and customers, fostering a growing and engaged network.

• Token Growth Potential: Given the high business value and innovative features of the project, there is a potential for the WML token to increase in value by more than ten times.

• Airdrops: The budget includes 2% of all tokens allocated for airdrops, providing an opportunity to earn free tokens and engage a wider audience in the project’s ecosystem.

WorkML.ai — highly profitable and low-risk feature-rich employment hub for investors, customers and annotators.

WorkML.ai transforms the landscape of the cryptocurrency market by delivering real value to businesses, investors, and a diverse group of users ranging from clients to data annotators. By moving past the speculative nature of token offerings, it solidifies a reliable revenue model based on service commissions, ensuring a consistent financial flow and anchoring the project's worth in practical benefits.

By meeting the tech industry's urgent demand for detailed data sets vital for AI system training, WorkML.ai significantly cuts the costs and time required for AI development. This enables wider implementation of AI technologies across various industries, providing high-quality data that improves the training and efficiency of neural networks.

Investing in WorkML.ai is more than just a financial decision; it represents a forward-looking partnership at the leading edge of AI innovation. It gives investors the opportunity to engage in a crucial initiative that promises significant returns and helps shape the future of technology.

Join the WorkML.ai Revolution

Explore the future of AI and blockchain at WorkML.ai. Our groundbreaking platform and WML token aim to revolutionize the training of AI models. Sign up for our newsletter to receive exclusive updates and get early information about our upcoming token sale.

We are open to new proposals and welcome collaboration (investor use-case). investors@workml.ai

Get Involved

Connect with us on our website and social media to participate in webinars and join our growing community. Your insights are vital to our collective success.

Website | LinkedIn | Telegram | Facebook | Instagram | YouTube | Twitter | Threads