Inspiration: The inspiration came from a conversation with my father. We talked about how many of Ethiopia’s historical books could one day vanish due to a growing lack of skilled transcribers. Even when transcribers do exist, manually writing an entire book on a keyboard is extremely time-consuming. Reading and preserving these works digitally should be easier and faster and that realization became the foundation for this project.
What it does: it is a voice-to-text transcriber, refine the transcribed text, export in PDF, TXT,DOCX,SRT, edit text.
How we built it: We built it on Google AI Studio .
Challenges we ran into: One of the main challenges we faced was the lack of clear idea from transcribers about what kind of platform they actually needed. Because of this, we designed the platform based on how we would want it to work if we were transcribers ourselves. Another major challenge was translating our ideas into effective prompts getting the AI to fully understand exactly what kind of picture/imagination we were drawing in our mind.
Accomplishments that we're proud of: This is my first hackathon that I have pushed my self to explore and see what other opportunities provide, to learn from other people and step out of my comfort zone . Working collaboratively with my brother and father, even though we live together we haven't worked on something together as a team before.
What we learned : Ideas are easy but implementation is hard. We use use our imagination to the fullest but never once realized that turning imagination to reality is far more challenging that it seem. Yet, at the the end of the day , our imagination is what keeps us moving forward.
What's next for Amharic Scribe AI: The next step for Amharic Scribe AI would be to have to fullest capacity to understand Amharic and to add new languages
Log in or sign up for Devpost to join the conversation.