Personal details

Adrian T. - Remote

Adrian T.

Timezone: Madrid (UTC+2)

Summary

I like to bring outside-the-box thinking to your problem, having worked in different technical fields includig scientific research, and having lived in 4 different countries, speaking 4 different languages.

Work Experience

Freelance iOS Consultant
adriantineo.com | Mar 2019 - Present
Git
Xcode
Unit Testing
Machine Learning
Continuous Integration
Swift
iOS
Objec
Craft native iOS apps as per clients needs
iOS Developer
Scout24 Switzerland, AG | Nov 2016 - Nov 2018
SQL
Git
Xcode
Unit Testing
Objective-C
Swift
Agile
I worked as an iOS developer for 2 years on the ImmoScout24 app, serving thousands of daily users in iPhone and iPad. During this time, the team accomplished many significant milestones, from releasing big features, a big redesign, bug fixing, and a big refactoring. In particular, I led a long-term effort to evolve a large legacy codebase in Objective-C to Swift, based on SOLID architectural patterns and best practices overall.

Personal Projects

Enabling iOS apps with machine learningIconOpenNewWindows
2019
Machine Learning
Swift
iOS
I was accepted to speak at AltConf Madrid 2019 (3rd June 2019) on the subject of enabling iOS apps with machine learning with CreateML. I conducted independent research on the possibilities of machine learning in iOS apps in challenging problems taken out of 3 high-paying Kaggle competitions: image classification, language processing (NLP), and numerical data prediction. I compared scientific algorithms based on common libraries such as Keras, TensorFlow, and XGBoost with CreateML, Apple's high-level tool to create, train, and evaluate machine learning models in the native CoreML format.
3D image segmentation machine learning pipeline
2019
Python
Linux
Machine Learning
PyTorch
I implemented a fully functional ML prototype pipeline for the segmentation of 3D images, based on a state-of-the-art solution: from image preprocessing, to training of deep neural network with GPU acceleration, translation of resulting embeddings, and interpretation of generated results.