A – N.A.O.C. — A Prompt Framework for Generative Image Models
Agent product for ZDES Events · 2026
CG production: models, look-dev, renders and films · 2025–2026
ORIS: 8 models, 28 variants, 212 images and 14 films — CG before the watches go into production
AI generalist and supervisor · 2025
AI supervisor and AI generalist · 2026
A 60–75 second manifesto film about a studio where live action meets AI. Moscow, winter, natural light, a touch of grain. One lead who doesn’t act, just exists in the city.
ZDES Studio pre-production tool · 2026
Storyboards as a Composition Lock in Generative Production
3D design for a Buro 24/7 special project · 2021
Film art direction · 2021
A Generative-AI Video Production Pipeline
Typed-Reference Composition: Every Reference Image Has One Job
Oval face, shoulder-length dark wavy hair, brown eyes, light tan, warm smile.
Morning at the gym, treadmill, water in hand
House party, laughter, all handheld
The couch after a long day, soft light
Sunny selfie, the brand’s product in frame
Character system for influencer marketing · 2026
Photogrammetry and CG production for a digital campaign in Russia and the CIS · 2022
VFX & CG · 2023
Generative production · 2025
Art director and CG supervisor · 2022–2023
Katy Perry · Viva Vision: a show on the world’s largest single video screen, about 30,000 people at the premiere
Art direction and CG/VFX for BBDO · 2021
Music Video · Art Direction · 2021
Music video · Post Tribe
Music video
Three orthogonal axes of a shot (Anchor, Optics, Chemistry) that make an image prompt debuggable.
Identity and brand preservation as a separable prompt architecture. The generative prompt describes this shot; the Lock Layer wraps it with an immutable structured constraint, replayed verbatim across every generation in a batch.
A prompt architecture for multi-input image generation: every reference carries an explicit typed role, and the prompt names inline which reference contributes which dimension. Stops silent-reference hallucination at its source.
A reference architecture: 6 stages from brief to video, 3 consistency mechanisms, and a dependency graph.
Four prompt architectures that hold a character and a style across 30 shots without LoRA or fine-tuning.
Hand-drawn-style storyboards aren't decoration in a generative-video pipeline — they're a cheap visual lock on composition. Cheap to produce, fast to reject, and they double as vision references for the final-render stage.
Send the task or idea as it is, even as a rough draft. We’ll think about what it could grow into.