Filip Vinković

I've been working solo on a full-stack meal-planning app focused on nutritional values.

It offers features such as a meal calendar, meal plans, shopping lists, a recipe database, and more, with all user data synced across devices. Nutritional values are taken from a third-party service and integrated into the app.

I built the back end using Kotlin (Ktor) and PostgreSQL, the web front end using React and Next.js, and mobile apps (Android and iOS) using Flutter.

I integrated AI (OpenAI API) to make recipe importing even smoother by matching user's input with foods from our food database.

As a member of an agile team consisting of four iOS and four Android developers, I implemented and maintained critical parts of the app, which has more than 500,000 downloads.

Some of the complex features I worked on include ad details, an in-app messenger, a dynamic form input engine, and more. I participated in scrum meetings, time estimates, and other related operations. We wanted an easy-to-maintain, testable codebase, so we used clean architecture, an MVP for the presentation layer, and Dagger for dependency injection. We composed unit tests with JUnit and UI tests with Espresso. Tests and static analysis tools such as Lint and Detekt were run on the GitLab CI server to prevent issues from being merged into the main branches.

Tools and Libraries: Android SDK, Java, Kotlin, RxJava, Retrofit, OkHttp, Dagger, Glide, Picasso, Facebook SDK, JUnit, Mockito, Espresso

As an Android developer in an Agile team, I created a white-label app that was published for ten different amusement parks in North America. Each of the apps was downloaded between 100,000 and 500,000 times.

The app downloaded a configuration from the API to modify the appearance of the app, depending on the park it was used for.

I implemented Bluetooth beacon detection with Bluetooth Low Energy API to trigger push notifications as the users walk by a certain POI in the park.

I worked with location services to display the user's location on an illustrated high-resolution map which utilized tiling to make pinch-to-zoom performant and reduce the device's memory usage.

Geofencing was implemented to detect when the user enters and exits the park.

Technologies and Tools: Android SDK, Bluetooth Low Energy, Java, Retrofit, DBFlow, JUnit, Mockito, Glide

I worked in a team of three Android developers on the development of an electricity consumption monitoring app.

The app was written in Kotlin and followed clean architecture and MVVM. Android Architecture Components were used, specifically ViewModel and LiveData. To ensure code quality, we used Detekt, Ktlint for static analysis, and also wrote JUnit tests.

We forked and modified a popular open-source chart library to display energy consumption with real-time data updates.

OAuth was implemented for authentication.

Technologies and Tools: Android SDK, Kotlin, RxJava, JUnit, Retrofit

This was an Android app that used Bluetooth Low Energy to connect to temperature, humidity, pressure, and light sensors to read measurements and publish them to an API.

When I joined the project, the codebase was in a rough state and my task was to improve reliability and implement new functionalities.

I fixed a number of concurrency issues, improved performance, and stability which resulted in fewer missing data points.

I also implemented data traffic throttling when the devices are connected to the cell network which reduced data usage costs by a fair amount.

Technologies and Tools: Android SDK, Bluetooth Low Energy, Java, Couchbase SDK

I created an Android app that allowed users to save song lyrics, add chords to them, and create playlists.

The app was written in Kotlin and followed clean architecture and MVP for the presentation layer. Unit tests were written with JUnit and Mockito.

I implemented a custom view for the optimal performance of rendering lyrics with music chords with support for touch gestures like scrolling and piching to change text size.

The lyrics view featured a pinch-to-zoom and automatic scrolling capabilities with configurable scrolling speed.

I used Realm for data storage and Retrofit for connecting to a RESTful API.

I closely worked with the back-end team to design a data synchronization system between multiple devices.

Technologies and Tools: Android SDK, Kotlin, Realm, Retrofit, JUnit, Mockito

A financial application for money transfer between people, depositing, and withdrawing from partner companies and telecom accounts.

The app was written in Kotlin, using the MVVM architectural pattern (Android architecture components—ViewModel/LiveData) in conjunction with dependency injection (Koin) to ensure maximum testability. Threading was done using Kotlin Coroutines, and unit tests were written using a combination of JUnit/Mockito. Builds run through a continuous integration pipeline on GitLab that includes static analysis checks (lint, detekt) and unit tests.

For this project, I implemented OAuth, push notifications, and social media sharing.

Per the client's requirement, I implemented fetching encrypted articles from the API and decrypted them in the app using the AES encryption algorithm

Technologies and Tools: Android SDK, Kotlin, Retrofit, DBFlow, JUnit, Espresso