# Morsekey — QSL Card Archive with AI-Powered Search

Year: 2022  
Type: Personal  
Category: web  
Page: https://eliachiarucci.com/projects/morsekey

A collection of letters exchanged by radio amateurs all around the globe. Search by image content or full-text search on OCR scans, powered by AWS Rekognition.

## Overview

Morsekey is a digital archive of **QSL cards**, the letters radio amateurs exchange to confirm contacts, collecting over 15,000 scanned cards from operators all around the globe, dating back to 1961.

Every uploaded scan lands in **AWS S3**, which triggers a **Lambda** function that runs the image through **AWS Rekognition** for both OCR and image-content labeling. The extracted text and labels are stored in **MongoDB**, while the scans themselves are served through CloudFront.

The **Next.js** front-end queries this index to offer three ways in: search by call sign, full-text search across the OCR'd card text, and search by image content (find every card with an elephant, an antenna, or a map on it). The result is a fully searchable, browsable historical record of decades of amateur radio correspondence.

## Key Features

- 15,000+ scanned QSL cards from radio amateurs worldwide
- AI-powered OCR full-text search across card scans
- Search by image content via AWS Rekognition labels
- Serverless ingestion — S3 uploads trigger Lambda analysis
- MongoDB search index with CloudFront-cached images

## Tech Stack

Next.js, React, MongoDB, AWS S3, AWS Lambda, AWS Rekognition, CloudFront, Vercel

## Links

- Website: https://www.morsekey.net/qsl