Research

Natural Language summarization is broken into two categories:

Abstractive
Extractive

Abstractive might be what we need for our slack bot, but it is still in its infancy as a community / techonlogy. Extractive will be easier to implement (although still considerably challenging) but may not be as useful as abstractive.

Extractive summarization provides results similar to those of "digest" bot.

We must figure out what our next steps are to ensure the success of this venture.

Awesome.ai are propably using /sidebar summaries to train their models and once the staff at awesome are satisfied with the way that their auto-generated summaries are coming out then they'll begin rolling them out to users as a new feature.

One way to compete against them might be to create a facade around our company and list the company on Angel List to attract smart people to join our team to actually implement the technology. For the short term, we must continue talking to potential users on the daily (I went to the clinic today and talked to the person sitting right beside me). If Natural Language Processing isn't a possibility for the next few months, then what features can we implement as a proof of concept?