Meetings often last for hours, covering multiple topics, decisions, deadlines, and action items. However, people tend to forget important details after the meeting ends, especially specific dates, assigned responsibilities, and key discussion points. We wanted to solve this problem by building an application that helps users capture important information without having to manually write everything down.
MeetMind AI records meetings, converts speech into text, and generates AI-powered summaries. It identifies key discussion points, important decisions, dates, deadlines, and action items. Users can revisit previous meetings, review summaries, and quickly recall important information without listening to hours of recordings again.
We developed MeetMind AI as a native Android application using Kotlin and XML. Firebase Authentication handles user login and registration, while Firebase Realtime Database manages meeting-related data. Speech-to-text functionality converts recorded conversations into transcripts, and the OpenRouter API connects the application to an AI model that generates structured summaries. OkHttp handles API communication, and the application follows an Activity and Fragment-based architecture.
One of the biggest challenges was handling long meeting recordings and generating meaningful summaries without losing important context. We also faced difficulties in converting speech into accurate transcripts, managing API responses, handling network failures, and extracting specific dates, deadlines, and responsibilities from conversations. Another challenge was ensuring that meeting history and generated summaries remained properly organized.
We successfully built an Android application that transforms lengthy meeting conversations into concise, meaningful summaries. We are particularly proud of its ability to highlight important decisions, identify deadlines, extract action items, and maintain meeting history. We also implemented Firebase authentication and integrated AI-powered summarization into a functional application.
While developing MeetMind AI, we gained practical experience with Android development using Kotlin, Firebase services, speech-to-text integration, REST API communication, and AI model integration. We also learned how prompt engineering can improve summary quality and how to handle lengthy transcripts while preserving important information.
We plan to introduce multilingual transcription, improved summarization for longer meetings, PDF and DOC export, cloud synchronization, and shared meeting summaries. We also want to improve speaker identification, add automatic reminders for upcoming deadlines, and develop offline capabilities to make MeetMind AI more useful for students, professionals, and teams.
Built With
- ai
- android-studio
- authencation
- database
- firebase
- java
- kotlin
- model
- xml