Marko Ratkovic
Software Developer
Marko is a software engineer with a master's degree in computer science and professional experience in C++, Python, and JavaScript ranging from web development to optimization problems, machine learning, and data science. He's detail-oriented with excellent communication skills, focused on meeting the client's requirements and delivering clean and efficient software solution both as an individual developer and as a member of a team.
Availability
Preferred Environment
Git, PyCharm, Vim Text Editor, Bash, Linux
The most amazing...
...project I've worked on is a tool for aligning error-prone sequencing data to the genome.
Work Experience
Software Engineer
Mireo
- Worked as a member of the navigation team developing a platform-agnostic car navigation software (Windows, WinCE, Android, iOS, Linux, and so on).
Data Scientist
HashCode
- Developed a machine-learning-based solution for predicting a telecommunication network's performance.
- Oversaw the data cleaning, feature extraction, and creation of a predictive model for early detection of network performance degradation.
R&D Engineer
Span
- Worked as part of the computer vision research and development team.
- Developed a Raspberry Pi-based solution for people detection from a video feed and estimated their age and gender.
- Constructed a dashboard in C# for Raspberry Pi-based people-detection project used for real-time reporting, deployed on Microsoft Azure platform.
- Worked on a type of multicamera eye-tracking software.
Research Officer — Intern
Genome Institute of Singapore, A*STAR
- Developed a tool for efficient protein database searching that allows biologists to detect distant relationships between proteins.
Experience
Low-cost Surveillance System
Technologies: Python, DynamoDB, AWS Lambda, AWS S3
Real Estate Scraper and Notifier
https://github.com/mratkovic/njuskalo-notifierTechnologies: Python, Scrapy, SQLite
MinCall | MinION End-to-end Convolutional Deep Learning Basecaller
https://arxiv.org/abs/1904.10337I wrote a paper shows that the accuracy of sequencing data is not only limited by sequencing technology, but also by the underlying software used for base calling. I also developed software that improves the accuracy of basecalled data as well as faster basecalling speeds over official MinION basecaller (Metrichor).
Technologies: Python, TensorFlow, Docker
Kaggle Competitions
https://www.kaggle.com/rmarkoThis competition centered around an image-segmentation problem, specifically involving the detection of nuclei in cell images.
• https://www.kaggle.com/c/data-science-bowl-2018
• Result: 69/3634 — Top 2%
• Technologies: Python, Keras, OpenCV
Kaggle Competition 02
Here, this competition involved a variation of the traveling salesman problem with a modified loss and 198,000 cities.
• https://www.kaggle.com/c/traveling-santa-2018-prime-paths
• Result: 42/1874 — Top 3%
• Technologies: C++
Skills
Languages
C++, Python, Bash, Java, JavaScript, SQL
Frameworks
Django, Boost
Tools
Git, Visual Studio, Vim Text Editor, PyCharm
Paradigms
Object-oriented Programming (OOP), Unit Testing, Design Patterns, REST, Concurrent Programming, Functional Programming
Libraries/APIs
OpenCV, Keras, Scikit-learn, TensorFlow
Platforms
Linux, Android, Docker, Heroku, Windows
Storage
PostgreSQL, Amazon DynamoDB, Redis, Amazon S3 (AWS S3)
Other
Data Visualization, Data Preprocessing, Machine Learning, Mender
Education
Master's Degree in Computer Science
University of Zagreb - Zageb, Croatia
Bachelor’s Degree in Computer Science
University of Zagreb - Zageb, Croatia