Sure. Besides spike sorting, there are analogous methods in other domains:
in optical physiology, you need to identify regions that correspond to neurons from the microscopy image sequence and extract the activity vs time for each of those regions. This is generally called "image segmentation". The most popular packages for this are suite2p and CaImAn. EXTRACT is also promising but it is MATLAB-based so we'd probably need to use our containerization tricks to get it integrated.
automated pose estimation is becoming increasingly common. This takes video of behaving animals as input and outputs pose over time. DeepLabCut and SLEAP are the most popular, and a newer option Lightning Pose is very promising. The tricky thing about this is it requires manual annotation ahead of time, which necessitates a user interface that would ideally be integrated with the UI. I know SLEAP is heading in this direction so it might be a good idea to touch base about this.
Dimensionality reduction techniques such as CEBRA would also be nice to have.
@bendichter, could you please give one or two examples of processing on the DANDI NWB files that are not spike sorting?