We need a signature filter for Basecamp email inputs, RTM email inputs, and others, where the software learns to recognize the “signature” elements of the inbound email and removes them from the “active data” portion of the message. It should also apply retroactively, since in the abscense of an industry standard signature demarcation (excepting “–” which seems to be the only standard) it would take 2+ repetitions for the AI to determine what constitutes the static portion of the inbound messages. Alternatively, some other form of marking the active data could be used, but that would be a less clean solution and require user training.
Perhaps a single, open-source component could be developed and implemented in many places.
