From 2D Artwork to
3D Molds — and Digital Assets
A multidisciplinary production workflow spanning 3D modeling, physical manufacturing, photography, and AI-assisted asset creation.
THE CLIENT
This professional bakery that creates personalized treats designed to add a fun, unique touch to any special event or milestone celebration.
THE TOOLS
-
Carveco - Creative Design & Manufacturing
-
Nano Banana - AI-Assisted Asset Exploration
-
Photoshop
MY ROLE & CONTRIBUTIONS
3D Artist & Imaging Specialist
3D Modeling · Toolpath Development · Production Iteration · Photography · AI-Assisted Asset Creation
THE TEAM & COLLABORATION
-
Nisha Kabra - 3D Artist & Imaging Specialist
-
Bogdan Larovoi - Owner & Project Direction
-
Erika Figueroa - Project Coordination
-
Olga Mikhailova - Head Baker & Production
THE CLIENT
This professional bakery that creates personalized treats designed to add a fun, unique touch to any special event or milestone celebration.
THE TOOLS
-
Carveco - Creative Design & Manufacturing
-
Nano Banana - AI-Assisted Asset Exploration
-
Photoshop
MY ROLE & CONTRIBUTIONS
3D Artist & Imaging Specialist
3D Modeling · Toolpath Development · Production Iteration · Photography · AI-Assisted Asset Creation
THE TEAM & COLLABORATION
-
Nisha Kabra - 3D Artist & Imaging Specialist
-
Bogdan Larovoi - Owner & Project Direction
-
Erika Figueroa - Project Coordination
-
Olga Mikhailova - Head Baker & Production
THE APPROACH
From approved artwork to production-ready form, I worked through the problem in three connected stages.
-
Build the 3D Relief: I translated approved 2D vector artwork into high-fidelity 3D relief models in Carveco, carefully controlling depth and detail. I worked closely with the kitchen production team to refine the models based on real-
world testing.
-
Engineer for the Mold: I developed optimized CNC toolpaths to translate the final 3D models into precise industrial embossing molds, balancing the visual fidelity of the artwork with the requirements of physical production.
-
Extend the Asset Pipeline: Once the physical products were produced, I developed a hybrid photography and AI-assisted workflow to overcome traditional studio and retouching constraints and create polished, e-commerce-ready digital assets.
The Outcome
50%+ reduction in baked-good waste
Iterative testing and production feedback helped reduce baked-good waste by more than 50%, while allowing the molds to retain the intended details of the original artwork.
Beyond the physical product, the project connected three traditionally separate stages of the workflow: creative artwork, physical manufacturing, and digital commerce. I helped carry that process from 2D artwork through 3D modeling and production, then extended the finished products into polished, e-commerce-ready digital assets.
THE CHALLENGE
Turning flat artwork into a physical product—without losing the details that made it work.
Preserve the Design: The approved 2D artwork needed to be translated into dimensional reliefs while maintaining its recognizable forms and fine details.
Engineer for Production: The 3D reliefs required precise draft angles and depths so the physical molds could cleanly release the dough without cracking, distortion, or loss of detail during high-temperature baking.
Rethink the Asset Workflow: Traditional product photography introduced practical constraints—including studio costs, lengthy retouching, and natural imperfections on the cookie surface. The project needed a faster, more repeatable way to create polished digital assets.
.png)
.png)
Initial stage relief test captured the core concept but exposed text legibility limits.
Digital prototyping allowed quick iterations without consuming raw materials.
Early-stage feedback text depth adjustments improved accuracy and a 50%+ reduction in baked-good waste in the production process.
Exploratory Asset Generation With AI:
