Beagleboard-gsoc project idea discussion

My name is Anirban Banik. I am interested to work under as a part of gsoc. Have some experience programming with python and c, and a basic knowledge of node.js. I searched through the ideas for gsoc and found the speak and spell idea very interesting. Going through the idea, I felt that this was intended for making a simple but modernized version of the toy. However, I am unaware as to what capabilities should the new device have.

Searching the open web, I felt that we are basically trying to build a speech recognition and processing application. However, I wish to add some more capabilities to the old version of the toy. The newer version should aim at all age groups of people, for the adults it can function as a device which evaluates his speaking abilities, and skills, correcting whenever he has misspelled a word…as a result It can be of much help preparing for various exams where english speaking fluency is checked, and can provide an advanced analysis of the preparation level of the student. Am currently concentrating on the english part only, though it can be extended to other languages also.

Can anyone help me out with this idea or it is irrelevant in this context? I currently don’t have a beaglebone so can I start my groundwork without it?
Or any issues or other things on which I can work during this period…would be obliged if somebody guides me.
Thanks in advance,