Scaling Synthetic Video Generation Without Cloud Bottlenecks | Case study | Liquid AI

The challenge

An automaker needed massive synthetic video generation for simulations—but cloud costs were exploding, and rendering was too slow for agile development.

Key Obstacles

• High cloud costs: Expensive GPU instances drained budgets
• Slow rendering: Delays stretched development timelines
• No scalability: Cloud-dependent workflows couldn’t keep up

Our solution

Liquid built a lightweight, local synthetic video model that ran directly on automotive hardware, eliminating cloud dependency.

The results

Faster, cheaper simulations at scale— no cloud needed.

• 2x faster rendering speeds
• 70% lower cloud costs
• Scaled synthetic data production in-house
• Gained competitive edge vs. cloud-reliant rivals