Lectro

Inspiration

Being students ourselves, we understand the struggle of trying to keep up in lecture, especially courses that cover complex topics. Our goal is to ensure that every student -including those with disabilities- to be able to fully keep up in class. We've all had those professor's that would just read off the slides or just go over material too fast and sometimes, if you are too busy taking notes, you won't be able to understand what you're actually writing. So by combining real time transcription and AI assisted note taking, we wanted to create a tool that allows everyone to follow along, regardless of how complex or fast the material being covered, allows you to understand the key concepts and have it stored.

What it does

Lectro is an AI-powered note taking assistant designed to help students backgrounds. It listens to lectures in real time and transcribes speech into text real time. -Utilizes AI to create notes in a bullet point format about key terms. -Automatically generates summaries and bullet points after class

How we built it

-Frontend: React + Vite + TailwindCSS -Speech Recognition: Web Speech API (Fallback of Whisper API) -AI Notes/Summaries: GPT-4o (with OpenAI API) -DataBase: FireStore

Challenges we ran into

-Getting the Web Speech API to consistently pick up and recognize niche words -Creating the backend AI logic for processing the transcript to generate accurate and precise notes -Coordinating and splitting tasks efficiently across group member

What we learned

-Full stack development -Collaborating with peers on a project -Using React hooks for managing the continuous audio streams -Building something that will make a difference in the real world

What's next for Lectro

-Implementing Key term Highlights and definitions -Adding an option for translating into different languages (for students of all languages or international/exchange student) -Developing an mobile app for more convenient usage --Stores the notes inside each persons personal account for future referencing

Built With

• firestore
• react
• tailwindcss
• vite
• webspeech