Project Info

Delulu Bot

This project did not submit a demo video on Devpost.

Inspiration

We wanted to create something which would bring joy to us and anyone else who tried out project. Since AI is all the craze these days, we thought it would be a fun twist to have a chatbot which is a bit different from the usual kind and responses with more sass/sarcasm and young people friendly slang.

What it does

The user records a voice message about anything they wish to share, ask about, or are just thinking of in general. Then our chatbot replied in kind, mirroring the kind of energy it detects. If the user is nervous, it reassures. If the user is confident, it boosts their ego and thinking to make them feel even better. Essentially taking on the role of a super supportive close friend.

How we built it

We first used React and Python to build a basis website for our chatbot. We then added functionality to record and store audio. After that, we made it so that the audio on submission would call our Hume.AI API. We had trained a model on Hume.ai to detect sentiment (nervous, confident, etc) so that we could use this to determine the mood that the AI should respond with. Once the sentiment of the audio file was analyzed, it was sent back though Together.AI where we used the returned sentiment and confidence score to formulate a response with emojis and right type of vibe.

Challenges we ran into

We faced a bit of uncertainty when it came to the implementation and understanding of the Hume.AI API implementation as it related to our site. We were not highly familar with React nor Python, so setting up the site also took quite a bit of effort on our end. Overall it was a push of our knowledge and experience.

Accomplishments we're proud of

Since this project was a push on our capabilities, we are insanely proud of being able to have a functioning product to the level that it is. It was so fulfilling to be able to watch our known knowledge increase and grow along with the codebase.

What we learned

We learned more about React, Python, and utilizing Ai models for our personal usecases, especially in the form of training them.

What's next

In the future we plan to improve the site UI/UX, add in functionality for negative thoughts being respoonsed to with either redirection to therapists or therapy bot, and improved responses maybe in the form of the chatbot sending voice messages back in response.

Analysis

Compare with all teams

View

Metric

Ayumad
3
Nimita Mishra
3

Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.

Technology

Found in codeClaimed only

CSSIn code
HTMLIn code
JavaScriptIn code
PythonIn code
ReactIn code

5 of 5 appear in the indexed code.

AI coding agents

No AI coding agent signals were found in this repository.

Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.

Codebase size

Source size

15 KB

Source files

Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.