10 years in CGI and computer graphics. I have worked as a CG Generalist, CG Supervisor and Art Director. 6 of those 10 years were spent in major Russian studios: Grad, Synticate, FibrFilm and CarbonCore. Another 4 years were freelance, where I led projects more independently: from breaking down the brief and pipeline to the final result.
My main focus is post-production for advertising and film. Over that time I have gone through different roles and production stages, so I understand the whole pipeline: from the idea, direction and visual solution to technical execution, compositing and final frame quality.
I shoot with FPV drones and understand the process not only as an operator, but also as someone who has assembled, tuned and maintained drones myself.
Most of my drone content was shot in Indonesia and Thailand. I have worked with well-known bloggers, hotels, villas and travel projects.
FPV is a way to guide viewer attention through speed, trajectory, scale and rhythm. Experience in CGI, editing and color helped me move quickly into trend-driven social videos: dynamic visual content that holds attention and gains large reach. My personal record is 116 million views on one video.
FPV became not just a separate direction for me, but a fast entry into the world of viral social media content. I started to understand viewer attention better: what catches people in the first seconds, how to hold rhythm and how to bring someone to the end of a video.
I create videos that involve people and keep their attention until the end. My CGI background helps me plan and assemble content that looks more premium than a typical social format while still staying clear for a broad audience.
Social media is also a platform for business. It is a way to quickly sell an idea, a place, a product, an atmosphere or an image through a short visual format.
On the wave of viral social content and in search of even stronger visual solutions, I started actively exploring neural networks. At first these were early mass-market tools such as Stable Diffusion 1.5 and early versions of MidJourney.
I worked with image and video generation, deepfake face replacement and real-time voice cloning. I set up a stream with a full identity replacement: voice through RVC, face through DeepFaceLive and a custom AI model trained on material generated with neural networks.
When generative models became strong enough for commercial production, I started making advertising videos built entirely with AI tools and closed paid generation models. In my work I use Kling, Seedance, Veo, Sora, Flux, Reve, Nano Banana, Seedream and other models.
Digital Products / Experiments
1. Qver is a startup created with my brother. It is designed to replace traditional 3D visualization with a one-button solution.
Qver already has more than 100 thousand registered users. Today it is a profitable business and a project that gave me hands-on experience in startups, product work and business development.
2. The educational project was an attempt to replace dry textbook information with engaging videos that use 3D graphics to clearly explain how the world works in physics.
I managed to make two videos about electricity. Back then everything had to be done manually: script, graphics, editing, sound and final assembly.
3. Experiments and development: in general, I approach work as an enthusiast and startup founder. I like coming up with new ideas, testing them quickly in practice, building prototypes, developing products and learning new tools.
Right now I am experimenting with vibe-coding, game prototypes, neural-network-based services and AI tools for development. This portfolio site was also built with that approach: I design the structure, visual logic and behavior, while AI helps write, check and evolve the code faster.
