CoreWeave needed to show what cloud computing actually makes possible — not through diagrams or dashboards, but through the raw physics of an F1 car tearing through smoke, light, and data.
Every frame in this film was built through a custom production pipeline — iPhone footage mapped in 3D, layered with generative AI, then tightened with hand-finished logos and compositing.
The process behind Coreweave / Aston Martin F1 unfolded across multiple nonlinear but interconnected workstreams — from building custom training datasets and filming miniature RC cars, to designing procedural UI systems and compositing final shots through layered AI pipelines.
Trained custom LoRAs on Aston Martin F1 data — liveries, aerodynamic forms, track environments — so the AI output felt native to the brand.
Built grayscale 3D animations in Cinema 4D as structural foundation, then textured and rendered through generative AI pipelines.
Shot miniature RC cars for dynamic motion — real physics, real light — then style-transferred the footage through AI.
iPhone footage of miniature RC cars, mapped in 3D, layered with custom-trained AI, finished with hand-placed logos. Every frame built from scratch.
RC cars were filmed on iPhone to capture real-world motion and light, then run through custom-trained models to transform the footage into photoreal Aston Martin F1 imagery.
Extensive visual exploration across multiple environments to define the film's aesthetic language before production.
The visual toolkit we considered to communicate cloud infrastructure as something felt, not explained.
Exploring how to overlay meaningful telemetry and data visualization onto AI-generated footage with no inherent physical reference points.
