Mimi AI Meeting Notes

Mimi is an AI meeting assistant published on the App Store and Google Play, designed to help users turn meetings and audio recordings into structured, searchable, and useful outputs.

The product started from a simple but recurring problem: long conversations often contain important decisions, ideas, tasks, and context, but most of that information becomes difficult to revisit after the meeting ends. We wanted to build a mobile-first assistant that makes meetings easier to record, understand, summarize, and ask questions about later.

Mimi supports both live meeting recording and uploaded audio files. After a recording or upload is completed, the app processes the audio into a transcript, generates a meeting summary, and creates a contextual AI chat experience where users can ask questions about the content of the meeting.

One of the core parts of the product is the meeting chat experience. Instead of only showing a static summary, Mimi allows users to interact with the meeting through an AI assistant that has access to the full meeting context. This makes it possible to ask questions such as what was decided, what tasks were mentioned, who said something, or what the main points of the conversation were.

The app also includes a transcript experience connected to playback. We designed this flow with inspiration from Spotify’s lyric display, allowing users to follow the meeting content in a more time-aware and readable way while revisiting the recording.

From a technical perspective, Mimi combines mobile development, backend systems, audio processing, transcription, summarization, contextual chat, and retrieval-augmented generation. The RAG layer is especially important because it allows the assistant to answer based on the actual meeting content instead of acting like a generic chatbot.

Mimi was built as a real market product, not only as a prototype. We implemented user authentication, freemium access, premium limits, mobile publishing flows, and production-oriented backend logic. Publishing the app helped us think more seriously about onboarding, user experience, pricing, retention, and how people interact with AI features inside a mobile product.

The current version focuses on individual users and small teams who want to record, summarize, and revisit their meetings more effectively. At the same time, we are currently developing a B2B pivot of the project, moving Mimi toward a workspace tool for enterprises where teams can manage shared meeting knowledge, collaborate around conversations, and use AI to access organizational context more effectively.

This project is important to me because it combines several areas I care about: mobile product development, AI interfaces, real user workflows, SaaS thinking, and the challenge of turning unstructured conversations into useful software experiences.

Links
👉 App Store