Technical Assessment

Example Answer 1:
The company has built proprietary natural language processing libraries and entity recognition to predict/classify the presence of different emotions within email text sent between employees of the same company or department.

Example Answer 2:
The company claims to be 2 months away from completing a prototype device that will collect electroencephalogram (EEG) data to process and classify electrical activity in the brain for use in controlling electronic devices such as cell phone or computer through cursor control and a virtual keyboard.

Example Answer 1:
Identifying emotions within email text is relatively straightforward with the use of off-the-shelf libraries that have been pre-trained. The company claims to have generated 10-20% higher accuracy than available through existing open source libraries for sentiment analysis and that the specific emotions they have trained their system to identify within emails are the most important for the managers of the companies they target as clients.

Example Answer 2:
While I'm not extremely well versed in this field, my understanding is that turning EEG signals into a reliable computer interface is extremely challenging because of the variation in the characteristics of the signals between people with different psychologies. They will need an extremely large dataset to be able to classify brain signals in the way they propose.

Example Answer 1:
Their approach appears to be working. In the future they may want to explore depp learning as a way of improving accuracy. This link from 2015 may be useful: https://www.microsoft.com/developerblog/2015/11/29/emotion-detection-and-recognition-from-text-using-deep-learning/
Example Answer 2:
I would suggest they add additional neuroscience expertise to the team. The founder I spoke with was not aware of some of the most recent advancements made in bio-metric signal processing that may be relevant to her technology. I knew about this advance despite it not being in my field.