Blobfish AI

Continuously generated, verifiable enterprise simulations for training AI agents at the edge of their capabilities. Simulation infrastructure for agent training AI models practice real enterprise workflows inside simulated tools, learn from exact feedback, and keep training as their capabilities improve. SC Built from first-hand experience. Six months building enterprise post-training environments. The useful tasks sit at the model’s capability boundary: difficult enough to expose failure, but possible enough to reward progress. The hard part is not building the first task set. It is building the next one.