E'm-pathy

• 

  Getting the user data to cater user-specific suggestions

• 

  The responsive AI chat dialog conversation for every query from the user

• 

  The mood scale helps the user select between range of moods that one experiences

• 

  Login/signup page for each user

Inspiration

Having no immediate solutions to sudden mental turmoil is no good place to be. Hence we have this solution cater to people's immediate need for therapeutic suggestions tailored according to individual needs and behaviours.

What it does

The product is backed with pretrained AI models that analyses user's current state of mind and the intensity of those emotions using NLP induced sentimental analysis and providing suggestive actions if the emotions are mildly negative and possible nearby help to reach out to if the emotions are extremely negative by any chance. The product also holds some extra features along with the MVP that cater a few specific immediate needs that aligns with the main functionality

How we built it

We have made use of a python backend that holds an api integration with a pipeline of AI models that goes in the following order NLTK, Sentimental analysis, Specifically catered NotebookLM.

Challenges we ran into

We still are figuring out the integration between AI apis, moving data between two different AI models appears to not be a straightforward task

Accomplishments that we're proud of

Getting the idea on paper and the initial shift from brainstorming to the implementation only sounded like a good drift, but we've mostly been lagging since then 😅

What we learned

AI integration doesnt appear that easy as its said to be. And the insane number of possibilities you can wrap the functionality across any AI solution does make it a lot complicated because everybody offers a equally powered AI right out of the box and its becoming even harder to judge.

What's next for E'm-pathy

Getting all the planned things into the code is still a hefty task to pull off. Having to deal with a lot of unknowns did create a lot of curiosity in us, so hoping that we still tend to work our way to complete it irrespective of the hackathon window

Built With

• golang
• googleai
• notebooklm
• python
• tailwind