Trajectory Uses Runloop to Scale Continual Learning Workloads

Trajectory, an AI company focused on agentic models, has announced it is leveraging Runloop's cloud infrastructure to scale its continual learning operations. The company is running over 10,000 burst-concurrent development environments (Devboxes) on Runloop's platform to continuously post-train its AI agents based on real product usage data at scale. This implementation allows Trajectory to adapt and improve its agentic models in real-time by incorporating feedback and learning from actual user interactions with their AI systems. The partnership demonstrates how cloud-native development environments are being used to support the intensive computational requirements of modern AI training workflows, particularly for companies implementing continual learning approaches where models are updated continuously rather than through traditional batch retraining cycles.