In 2014, Oleg Yusupov founded Phygitalism, a company focused on visual content development. To expedite the creation of such materials, Oleg’s team developed an AI tool, Phygital+, which later evolved into a startup that secured grants from major IT corporations. Oleg shared the journey of Phygital+, from its inception to development, and future business plans.
Tell us about Phygital+ and the market needs you aimed to address with this company. How did the idea come about?
Phygital+ is a workspace for designers and visual effects creators, taking the form of a no-code AI editor for producing digital content using over 30 popular AI tools for graphic work. Phygital+ uses a unified interface, allowing designers to work simultaneously with popular neural networks like Stable Diffusion, MidJorney, DALL-E 3, Kandinsky, and ChatGPT, and connect them seamlessly. The project was started in 2022, participated in American accelerators, and won grants from Open AI and Amazon.
I always wanted to create an interesting, high-tech product and, thus, chose visual content creation technologies (AI, XR, CG). In 2014, I opened my visual solutions studio, Phygitalism, but by 2018-2019, I realized that AR/VR technologies face significant barriers that hinder their mass adoption:
- Hardware: Powerful enough equipment is very expensive for mass use, and many ideas can’t be implemented due to such technical limitations.
- Internet speed: AR/VR requires enormous amounts of power and speed, which is not readily available everywhere.
- Psychological factor: AR/VR content consumption tools, like VR glasses, significantly increase the cognitive load of people, and not everyone is ready for this.
- Content: The demand for AR/VR content is still too low, making it expensive to produce for such a small market.
These factors led me to the idea that in order to popularize AR/VR, we first needed a product that simplifies and reduces the cost of creating augmented reality content. My team and I began thinking about how to make development more efficient, faster, and cheaper, and came to the realization that AI was essential.
Within our studio, CG designers worked alongside data scientists, sharing experiences and combining CG and ML technologies. By the end of 2021, this crystallized into Phygital+—we developed a 3D-ML technology for rapid 3D content generation. Moreover, based on this, in collaboration with Microsoft, we launched a master’s program at the Moscow Aviation Institute. Initially, it was an internal tool for our studio, but after recognizing its potential to help many graphic designers, we decided to promote it to a wider audience.
What principles and values underpin Phygital+?
When I created Phygitalism, I adhered to the principles of a teal organization—self-management, wholeness, and the absence of hierarchy. While developing Phygital+, we kept these principles as a guideline, but since a startup needed structure in its early stages, we added standard business processes like OKR methodology and Agile.
Our mission is quite global: to help civilization find balance with the environment. This can be influenced through three levels—ecology, economy, and mindset. Phygital+ focuses on working with the mindset.
Some believe AI hinders creativity by simply generating images for the artist. But this is not true; when photography emerged, many painters feared they would lose their jobs. However, technological progress spawned many new art movements: futurism, modernism, impressionism. Artists realized that realism no longer surprised or impressed anyone, and so rethought their creativity.
The same goes for artificial intelligence, and those who do not wish to develop in the new reality may face difficulties. But we believe in the concept of augmented intelligence—when humans and machines work together, which allows specialists to maximize their capabilities and earn more. I think everyone will use AI in the future because it significantly simplifies life and makes work more productive.
AI specifically helps artists by enabling them to generate many visual concepts in seconds, from which the artist can choose the most interesting ones to refine. This greatly accelerates the creative process.
How long did it take to create Phygital+, and how did grants from Microsoft, Amazon Web Services, and Google Startups help you?
We started internal tests for Phygital+ in 2019 and realized its potential in 2021, working on it intensively. By September 2022, we released the first alpha version.
I believe grants are an excellent tool for product development, and all startups should pursue them. For instance, I personally applied for AWS and Google grants. The AWS grant greatly assisted us with cloud computing. An AI startup requires high GPU power—AWS provided us with a $125,000 credit, which we used on cloud tools for the platform, speeding up development.
The Microsoft grant covered costs for cloud processors (CPU) needed for databases, hosting, and other routine processes. Additionally, through Microsoft, we received an OpenAI grant to use ChatGPT technology via API. Microsoft found us on their own as a company promoting phygital technologies.
Google provides startups with grants up to $200,000, depending on the level in the acceleration program. So far, we have received $100,000, which we used for applied tasks necessary for project development.
Both AWS and Microsoft, and Google, support startups through expert consultations, technical lectures, and other events.
We are also part of the NVIDIA Inception Program, which allows us to use NVIDIA technologies and participate in negotiations with venture investors. The company invited us to join this program. From NVIDIA, we received additional AWS credits and are negotiating to speak at their conferences and webinars.
What did participating in the Fuelarts + Tezos accelerator give you? Was it challenging to pass the selection?
The selection process had several stages. First, we had to submit an application, then pass an interview with a technical specialist, and finally wait for their decision. I think the interview was slightly more challenging for us than for other startups because our idea was not very clear at the application stage.
The accelerator is primarily focused on Web3 technologies since Tezos is a Web3 protocol. Therefore, we had to demonstrate how AI could be useful in this context.
Overall, the main benefit of participating in the accelerator was understanding how Web3 protocols work and how we can optimize them for decentralized computing. During the accelerator, we attended lectures, communicated with technical specialists from Tezos, and learned a lot.
Tell us about your approaches to team formation—what do you prioritize, and what strategies do you use to attract and retain talented specialists?
We created a company with a strong R&D focus, always engaging in scientific or scientific-adjacent research, or fields that are not yet widely explored in modern business.
Our work involves many different disciplines, so our team members should be interested not just in performing standard tasks but in working at the intersection of specializations—CG, AI, and XR.
When hiring specialists, we focus on selecting A-Players, and in our work, we use the OKR methodology. However, those who stay with us long-term are only those who share our principles. Regarding compensation, as a startup, we developed a stock option program and included almost everyone working with us.
Currently, 10 people are working on the product. Of course, this is not enough for large-scale development. To grow the team, we attracted investment from a Cypriot company through a SAFE instrument: the investor gives us money, which will convert into company shares in the next round. However, I don’t see the point in excessively expanding the team until we reach Series A funding.
We have great synergy with our investor, and we plan to spend the money, including on marketing, to achieve high revenue indicators and attract a new round at a higher valuation.
What future trends in AI do you foresee, and how might they impact your business?
I see three stages in the development of AI for business:
- Interaction between a specialist and AI: A designer receives a task and completes it with the help of artificial intelligence.
- Interaction between a company and AI: The company “hires” an AI agent who “manages” AI designers.
- Interaction between a company and an AI studio: Here, artificial intelligence fully handles the design department’s tasks—from idea development to the final concept.
We are currently at the first stage, but AI agents are already emerging, though further development is limited by energy consumption and neural network architecture. I think the next stages will be closely tied to technical breakthroughs in hardware and decentralized computing—this is where the integration of Web3 and AI comes in.
Legal issues are also becoming more pressing. For example, there are currently claims against Midjourney, Stable Diffusion, and ChatGPT for using copyrighted data. By applying Web3 in AI tools, we can track on whose materials the content was generated and, thus, respect copyright.
I build my business and plans considering technological advancements. With a decade of experience, I understand where the industry is heading, and all its trends are included in my roadmap.