# I tried deep research to write a cognitive science project - surprisingly smooth. Has anyone else used subjust-specific AI agents for big academic projects? How was it?