The problem LivStory solves

• Helps the family spend quality time together in these lonely times of the corona virus.
• Solves the short attention span problem.
• More comfortable even for less tech-savvy users like grandparents.
• Gives the user the flexibility to speak any story in any manner, and we provide the sound effects, freeing them from a limited number of stores which a platform like google playbooks provides.
• Educative Sounds help for better imaginative purposes
• Increases cognitive intelligence

Challenges we ran into

1. We had to build a database of sounds since we did not have any database as such available which suited our needs, then we converted the sounds to Bayes-64.
2. Deploying the text rank model: Since the NLP model we used had to be downloaded first, then only we could use it. This actually worked on localhost but when we tried to deploy it we faced errors regarding the installation of the model. We thought about various ways and then finally came through pickling the model as well as the to-words after downloading them locally and accessing them via relative address whenever the model API was executed.
3. Warming up containers, as we used Azure for deployment we faced issues regarding the latency as we deployed it on free tier first, we later figured out by switching to a higher tier we could keep this app working with very very low latency.