Creative Technologist • Virtual Production • Procedural & AI Workflows

I design production‑ready workflows that blend classic craft with modern AI, always with transparent credits, robust data ethics, and reproducible results.

AI generated A-pose image with Runway

What I Do

  • R&D for real‑time and offline pipelines (Unreal Engine, Houdini SideFX, Cinema 4D)

  • AI‑assisted ideation & previz (Runway generation, diffusion upscales, voice & timing)

  • Look‑dev & simulation (Gaussian Splats, lighting studies, Niagara FX)

  • Photogrammetry & asset prep (RealityCapture, ZBrush cleanup, Substance 3D Painter textures)

  • Virtual Production environments (scene optimization, lensing tests, tracking prep)

What I Do

Principles:

  • Human‑led, tool‑assisted. AI supports craft; it doesn’t replace it.

  • Clear authorship. Roles and tools are disclosed for each project.

  • Ethical data. No training on private datasets without permission; brand/IP respected.

  • Reproducible workflow. I keep node‑based graphs, settings, and renders for QA.

Process:

  1. Brief & constraints → goals, references, delivery specs

  2. Method selection → when to use AI vs. traditional CG; risk/benefit

  3. Prototype → small tests; quality gates; timing & cost check

  4. Build → assets, lighting, FX; performance targets (fps, polycount, VRAM)

  5. Review → color, continuity, narrative clarity; legal/brand checks

  6. Deliver → plates, edit, scene files; change‑log and reproducibility notes

Toolchain:

Unreal Engine • Houdini SideFX (Solaris/LOPs) • Cinema 4D + Redshift • RealityCapture (AI assisting 3D modeling) • Wrap (AI retopology) • ZBrush • Substance 3D Painter • Runway AI • Higgsfield AI • PostShot (Gaussian Splatting) • Adobe CC

AI 3D modeling “Thumbelina”:

→

AI generated meshes with Tripo AI

→

ZBrush resculpting in progress

Reimagining the gems of New York City through the lens of Stable Diffusion

AI in my artisric projects:

ZINE: MARS, BABY

Mars, Baby is an AI-generated zine created with RunwayML. The project investigates speculative futures of human life on Mars through the lensof birth and early childhood. Set in the year 5025, it explores the idea that incubators have evolved beyond sustaining life to producing lullabies — blending technology, care, and cultural imagination in an extraterrestrial context.

you can learn more about this project here(Zine:Mars, Baby).

PIXEL VR GAME PROTOTYPE: “A BRIGHT MOON ON A WINDLESS NIGHT”(in progress)

image to video

This project is a simple yet atmospheric VR game developed in Unity with vibe coding. It recreates a Kazakh aul(village) in pixel-art style, prioritizing mood and immersion over complex mechanics.

you can learn more about this game here(Notion-vr-pixel-game).

text to image

GAUSSIAN SPLATTING FOR VOUGE FASHION VISUALS:

Gaussian Splatting experiment I created for Vogue — pushing fashion imagery into new dimensional language. Learn more about Gaussian Splatting here(surfsplatting.github.io). The custom library I used to create these visuals can be found here(Nargiza’s custom GSplats library), the process is showcased here.

AI and NYU Global Village 2025:

I created a series of sticker designs featuring iconic Kazakhstan landscapes for the NYU SPS Global Village 2025 using Runway and Adobe Illustrator.

AI visual “Objects in mirror are closer than they appear”:

I generated the base photo in Runway, then used Nano Banana to clean specific parts of the image. After that, I brought the cleaned image back into Runway to generate the final animated video.