Ex-Google/Apple researchers launch Trajectory to enable continual AI learning
startup trajectory raised $15m to build a platform enabling ai models to learn continuously from real-time user interactions, addressing the static nature of current foundation models.
trajectory, a startup founded by alumni of google deepmind, apple, and openai, raised $15 million in seed funding led by conviction to solve ai's inability to learn continuously. unlike current models that remain static after training, trajectory's platform uses real-world interaction data to frequently post-train open-source models, claiming this approach beats frontier models on narrow tasks. initial customer decagon uses the system to retrain ai support agents weekly based on failed interactions, such as complex return requests, marking a shift away from off-the-shelf api dependence.