Hi Michael, currently we are using Twitter API as it integrates well with IBM Watson. We believe if other social media APIs allow the same access as Twitter API, we will be able to use it for our project.

Hi Inian, we are using the Twitter API so what we are assessing currently is only public info. So we probably wouldn't be able to assess the location or private tweets. But we understand it's a valid concern and we certainly would tread carefully.

Hi Yoav, thanks for the wonderful suggestion. We would definitely look into it.

Hi Martin, currently many of such cases go unnoticed. We were hoping that with our app, those tweets that might previously have gone unnoticed, could be picked up and people could help them. Even if it is just one life saved, it is still a valuable idea to us. For the demo, we used the keyword "sad" but of course for real use case, more research would have to be done before we can arrive at more suitable word choices.

Hi Anson, thank you for the suggestions. We were hoping that our app could be useful for relevant international agencies (In Singapore, we have the Ministry of Social and Family Development & Samaritans of Singapore). So the onus is actually on the organisations to reach out to them (once identified, they would be able to privately message them), to provide counsel before tragedy really happens.

Explain more the potential of the solution. Which keywords are you focusing on? :)

Interesting concept that can be evolved to general detraction of "mood" according to activities on Social Networks.

This is a 2 sided market where one particular market you can't really target as paying customers - "the suiciders". This also needs to take into account that the "suiciders" want to be helped. If the paying customers not seeing results ..it maybe very hard to scale. Good try!

Very important problem. Privacy concerns for users? Location info, private tweets, more source of social input..

Usecase is limited to one platform and not clear if it can be used across multiple platforms.