Scaled Cognition
Builds action-prediction AI models for enterprise CX, delivering deterministic, policy-compliant agents that don't hallucinate.
Publisher details
Scaled Cognition is a vertical AI lab building models purpose-built for enterprise customer experience (CX), rather than adapting general-purpose LLMs to the task.
Its flagship model, APT-1 (Agentic Pretrained Transformer), is designed to optimize around action prediction rather than token or language prediction, a departure from how conventional LLMs are trained. The model is trained on a fully synthetic data pipeline built specifically to pair conversations with the actions that should follow from them, and it's further refined through agent-to-agent self-play, a reinforcement learning approach similar to techniques used to master games like Chess and Go.
