Matija Stepanic

Customized the FlutterFlow file upload feature to integrate with Firebase Storage SDK. I used UploadTask Stream to monitor upload progress and optimized memory by handling files as streams. I also added the ability to get original file names and implemented a drag-and-drop feature for file uploads using the super_drag_and_drop library from pub.dev.

This project includes a set of reusable Custom FlutterFlow widgets and actions for barcode generation; it depends on the open-sourced package `barcode_widget` from the pub.dev. Many types of barcodes could be generated and shared as an image outside the application. Also, the project includes implementation from `ff-encrypt` (another open-source FF project I created), which unlocks the feature for the creation of barcodes with encrypted data with the Salsa20 Cryptography Algorithm. The primary usage of that set of functions is to share data with encrypted QR codes.

A set of FlutterFlow Custom Actions for simple string encryption and decryption based on the Salsa20 Cryptography Algorithm. This project's main dependency is the `encrypt` package from pub.dev. The project includes multiple reusable functions and a Demo UI for testing all available custom functions.

I integrated the blinkid_flutter package with FlutterFlow, which wraps native iOS and Android identity document scanning SDKs. I developed a custom FlutterFlow widget that uses device cameras to read identity documents and returns structured data. This data can be processed in FlutterFlow Action Chains based on custom business logic (possible until August 2024). This solution has been implemented in 1,119 different FlutterFlow projects.

I developed a CLI tool that extends Flutter CLI functionality. When a push is made to GitHub from FlutterFlow, a webhook notifies a local Node.js listener. After verifying the Git commit signature, local Bash scripts integrate the changes into the local working Git branch. Unix signals are then used to trigger hot reloads across one or more running Flutter processes.

I developed an iOS and Android app for The Birth Deck Cards using FlutterFlow. The app, built with Flutter, integrates Firebase and Rowy for data management. Users can interact with cards, view related videos, and make in-app purchases via the RevenueCat SDK.

A Node.js serverless back end consisting of several Firebase Cloud Functions with integration to a few external services for media management. The photo booth WPF application is integrated with the official Canon SDK to take photos from Canon EOS DSLR cameras and print photos taken in photo booths or mobile apps. I wrote the browser PWA in Angular and Flutter Android, and iOS apps are currently in private alpha. The plan is to open source all parts of the system by licensing photo booth WPF applications.

A Node.js back end written in NestJS and Parse Server with Firebase integration. The browser PWA application is based on Angular, and the native iOS and Android mobile applications were written in Swift and Kotlin. For marketing and sales purposes, I integrated the project with a few external services for automated data transfer between production data and internal reporting tools. As part of ScanShop open source, it was contributed to the Parse Dashboard project on GitHub.

A standalone web component for integration to any web application with a generic UI for accessing a device camera, uploading photos, and sending them to the Microblink Cloud-based OCR engine. As a sub-project of microblink-js, I created the Angular application with a desktop-to-mobile feature to start identity document scanning on the desktop and use a native mobile camera with higher resolution for better document scanning. I used Firebase and Cloud Firestore as a database for exchanging data between server-side and client-side components.

A Node.js CLI tool for converting Parse Server schema format to DBML for easy visualization of Parse classes and relations through external tools, dbdiagram.io and dbdocs.io. It's simple to create Parse classes and relations in the Parse Dashboard application, all data is stored in a MongoDB collection called _SCHEMA, and this CLI converts MongoDB JSON documents for DBML.

A Node.js-based back end on top of Parse Server with a SPA front end in AngularJS. SaveTime is a ticketing counter application in the cloud for handling waiting queues by providing a number that can be tracked remotely from the web application. The waiting line resolver has a simple dashboard for taking the next ticket to process the related person as part of their business service. The waiting line could be anywhere—at the bank, hair salon, doctor, etc.

An Angular-based environment with Scully for building static landing websites with headless content editing by Flamelink CMS (with data-owning in a protected Firebase project) with a seamless triggering site build and static HTML deployment on Netlify from Flamelink via a Cloud Firestore onCreate trigger. Since 2020, several static websites have been created from that private GitHub repository for internal projects. The plan is to document and open source it live in the next few months.

A Node.js back end based on the Sails.js framework with MongoDB deployed to Heroku and MongoDB Atlas. The back end is a proxy with an in-memory cache for transforming the content structure from the WordPress API on the site and preparing to display it in Android and iOS applications created with Xamarin.Forms in C#.