model inference

Crusoe describes itself as the AI factory company, designing, building, and operating infrastructure for artificial intelligence. It offers Crusoe Cloud, an AI cloud platform with NVIDIA and AMD GPUs, managed Kubernetes and Slurm, serverless fine-tuning, and managed inference, plus modular Crusoe Spark data centers and an energy-first approach using wind, solar, hydropower, geothermal, natural gas, and carbon capture.

AI cloud computingData center design and constructionenergy generationfine-tuningliquid coolingmodel training

Cumulus Labs builds the compute, model runtime, and application infrastructure a business needs to own its AI. Its platform includes Paladin for GPU capacity and orchestration, Ion for model inference and GPU kernels, and Talos for agents, routing, evaluation, and fine-tuning. Customers can adopt each product independently or connect the full stack, deploying on Cumulus GPUs or their own infrastructure.

agent executionevaluationfine-tuningGPU orchestrationmemory managementmodel-specific kernelsmodel swappingworkload placement and rebalancing

Geolava builds a world model for the built environment called Terra, an intelligence layer that reasons over how the physical world changes and predicts consequential physical change before it becomes physical failure. Its causal engine detects failure signals beyond standard inspections across physical assets such as buildings, bridges, and airstrips.

causal enginelive state embeddingspredictive failure analysisprivate deploymentworld modeling