Inspiration
Have you ever experienced panic due to inadequate preparation for public speaking occasions? We strongly believe in the saying that practice makes perfect. With this belief, our team aims to create a platform that empowers users to improve their public speaking skills through dedicated practice sessions and tools. Our goal is not only to provide a space for users to practice but also offer customized solutions for improvement as needed.
What it does
PitchMe is a tool created to enhance students' presentation abilities using cutting-edge AI technologies. By employing AI camera vision, it can recognize and address improper body posture during presentations. The app also offers Real-Time Audio Speech to Text functionality for converting speech into text, as well as NLP for improving clarity through paraphrasing the content. Furthermore, PitchMe evaluates the presenter's tone of speech to give feedback on emotional and persuasive aspects of their presentation.
How we built it
We used Visual Studio Code as the platform to develop our website and Jupyter Notebook to train our models. Python was the main language for developing our artificial intelligence, along with libraries such as MediaPipe, OpenCV, Librosa, NLTK and more. For the website development, we utilized Python along with the Flask framework for back-end and integrated HTML and CSS for the front-end.
Challenges we ran into
During the implementation phase, our team encountered challenges, such as difficulty accessing most Generative AI API keys available online, as most of them required a purchase for usage. The search for a freely accessible API proved to be time-consuming. Additionally, the competition, spanning over four days, demanded our team to navigate ideation, prototyping, and evaluation processes within this tight timeframe, posing a significant challenge to achieving our goals.
In alignment with the user-centered design philosophy inspired by Don Norman, the godfather of 'User-Experience' concepts, we aimed not just to design a product for someone but to design it with them. To gain valuable insights, we conducted surveys within the limited span of two days to understand people's presentation experiences and needs. Meeting the target for collecting sufficient survey data within this short timeframe proved to be a challenging yet crucial aspect of our project.
Accomplishments that we're proud of
With our team being dynamic, collaboration was seamless, and we were able to work together as a team. We managed to implement all of the development tasks we thought of adding more features along the way, resulting in a fully functional and user-friendly tool for presenters to enhance their presentation skills. The output of our AI algorithms met expectations, and completing all features within this time frame was a significant accomplishment we are proud of.
What we learned
We learned about various AI techniques and their applications for detecting, paraphrasing texts, and analyzing tone. We also explored the different methods we could use for real-time audio transcription.
What's next for Kopi O Kosong - PitchMe
Participating in this competition has enabled us to complete the initial version of our application prototype. However, our journey doesn't end here; we are determined to further refine our idea and transform it into a viable business.
If you are interested in supporting our venture and would like to contribute to the startup phase, please feel free to reach out to us via email or phone call. Your support will be instrumental in helping us turn our vision into a reality. 😎
Log in or sign up for Devpost to join the conversation.