RecapAI: Automated Audio Insight Platform by Rabyaan KhanRecapAI: Automated Audio Insight Platform by Rabyaan Khan

RecapAI: Automated Audio Insight Platform

Rabyaan Khan

Rabyaan Khan

The Context

Great ideas happen in the room, but they often die the moment the meeting ends. Manual note-taking is distracting, and traditional voice memos are just black boxes of unsearchable audio. I built RecapAI to bridge the physical-to-digital gap, creating a pipeline that listens, transcribes, and summarizes conversations so teams can focus on talking instead of typing.

The Challenge: The Disconnected Workflow

Most transcription tools require you to actively manage software while trying to run a meeting. If you use physical microphones, getting that audio into an AI processing engine is usually a clunky, manual transfer process.
The goal was to build a true "set-it-and-forget-it" system where the physical hardware in the room and the AI cloud operate as a single, seamless unit.

The Solution: An Edge-to-Insight Pipeline

I designed RecapAI as an end-to-end ecosystem, ensuring that the data flowed effortlessly from the real world to the web dashboard. The architecture is broken down into three core layers:
The Edge (Hardware): A Raspberry Pi acts as the physical ear in the room. It handles reliable, high-fidelity audio capture and automated, secure uploading to the cloud.
The Brain (Backend & AI): A high-speed FastAPI backend orchestrates the heavy lifting. It pushes the audio through Speech-to-Text models and leverages Large Language Models (LLMs) to filter out the noise and extract the core context, structure, and action items.
The Interface (Web App): A Next.js dashboard where users consume the data. Instead of presenting a massive wall of text, the UI delivers clean, scannable knowledge cards.

My Role & The Tech Stack

I handled the end-to-end system architecture. The core challenge was ensuring a stable, modular data flow from edge hardware to user-facing outputs.
Hardware / IoT: Raspberry Pi
Frontend: Next.js
Backend: FastAPI
Intelligence: Custom LLM integrations
Database: MongoDB
The Outcome
RecapAI demonstrates that capturing knowledge shouldn't add friction to your day. By combining IoT hardware with cloud-based AI, this system successfully turns unstructured physical conversations into a searchable digital archive—ensuring that important information is never lost to distraction.
Like this project

Posted Mar 8, 2026

RecapAI is an intelligent audio transcription and summarization platform designed to capture spoken conversations and turn them into structured knowledge.