Marcel is available for hire
Hire MarcelMarcel Zięba
Verified Expert in Engineering
Data Structures Developer
Location
Warsaw, Poland
Toptal Member Since
March 8, 2017
Marcel has over ten years of experience developing fast and highly scalable applications. He has worked for big companies, including Google, but also for small and successful startups. He is a fast learner and he loves working with cutting-edge technologies.
Portfolio
AiFi
Swift, iOS, Python, PyTorch, Computer Vision, Deep Learning
Purple Squirrel
PostgreSQL, Node.js
Google
JavaScript, C++, MapReduce, Java
Experience
Availability
Part-time
Preferred Environment
Git, Python
The most amazing...
...program I've developed is used by more than hundred millions users every day.
Work Experience
Staff Research Engineer
2017 - 2018
AiFi
- Designed and trained fast and accurate deep learning models for pose estimation and human segmentation.
- Created a framework for running PyTorch models in real time (iOS and Android).
Technologies: Swift, iOS, Python, PyTorch, Computer Vision, Deep Learning
Software Developer
2016 - 2017
Purple Squirrel
- Developed back-end service responsible for transactional communication: email and SMS.
- Developed a service responsible for payments and refunds.
- Created multiple conference rooms for our users using Twilio.
Technologies: PostgreSQL, Node.js
Software Engineering Intern
2016 - 2016
- Developed a machine learning model that recognizes the main actors on photos with more than 80% of accuracy.
- Extracted and filtered photos for crowdsourcing.
- Developed an application for a crowdsourcing platform.
- Worked with crowdsourcing workers in order to improve task instructions.
- Developed multiple MapReduce jobs that extracted features from photos and trained machine learning models.
- Evaluated a machine learning model on an unseen set of photos.
Technologies: JavaScript, C++, MapReduce, Java
Software Engineering Intern
2015 - 2015
Imo.im
- Improved the probability of delivering a SMS by 3% and lowered cost of sending a SMS by 8%.
- Developed an in-app that handles a camera in Android application.
- Added a native VP8 codec for video-encoding in an Android application.
Technologies: C++, Python, Android, Java
Software Engineering Intern
2014 - 2014
- Developed an internal tool for a security team.
- Developed a highly scalable back-end for processing terabytes of network data.
- Created a user friendly web interface.
- Ensured the high security of stored data.
Technologies: JavaScript, Google App Engine, Python
Software Developer
2013 - 2013
Filmaster (acquired by Samba TV)
- Created a new movie recommendation algorithm that performed 10% better than the previous one.
- Developed a 1-page JavaScript application for the film festival.
- Created a Facebook quiz application.
- Developed multiple tests to ensure the correctness of a recommendation API.
Technologies: JavaScript, PostgreSQL, Django, Python
Software Developer
2012 - 2013
Future Processing
- Developed high performance algorithms for computer vision.
- Improved the stability of floating point computations.
- Created an algorithm for optical character recognition. The algorithm used an artificial neural network.
- Worked on internal tools that helped developers to create new functionalities.
- Helped to create a custom application for one of the clients.
Technologies: C#, C++
Experience
Imo.im Android App
https://play.google.com/store/apps/details?id=com.imo.android.imoim&hl=enImo.im is an instant messenger used by more than hundreds of millions of people every day.
I developed the front-end and back-end logic for creating video messages. I also implemented the handling of the newest VP8 codec. I used Java, C++, and Python.
I developed the front-end and back-end logic for creating video messages. I also implemented the handling of the newest VP8 codec. I used Java, C++, and Python.
Filmaster
http://filmaster.comFilmaster provides cloud-based recommendation technologies and solutions for cable providers, VODs, IPTV platforms, and cinema chains.
I developed the cutting-edge recommendation engine. The result was that the performance of the custom-designed model was much better than the standard machine learning models.
I developed the cutting-edge recommendation engine. The result was that the performance of the custom-designed model was much better than the standard machine learning models.
Adaptive Vision Studio
https://www.adaptive-vision.com/en/Adaptive Vision Studio is the most powerful machine vision software that does not require programming skills. It is based on dataflow and comes with a comprehensive set of image analysis filters optimized for industrial quality inspection.
I developed multiple computer vision algorithms including optical character recognition. I created an end-to-end solution for automatically reading personal data from a Polish identity card.
I developed multiple computer vision algorithms including optical character recognition. I created an end-to-end solution for automatically reading personal data from a Polish identity card.
Wonderlens
https://itunes.apple.com/us/app/wonderlens/id1267396805?mt=8Wonderlens is an iOS application that is able to segment humans in real time. I was responsible for designing, training, and running a deep-learning model on iOS devices.
Skills
Languages
Python, C#, Swift, Java, JavaScript, Haskell, C++
Frameworks
Swagger, Django, Bootstrap
Tools
Git, GitHub, Android Studio, Slack
Other
Machine Learning, Algorithms, Data Structures, Computer Vision, Security, Deep Learning
Paradigms
Data Science, Agile, Object-oriented Programming (OOP), Test-driven Development (TDD), MapReduce
Libraries/APIs
PyTorch, Node.js, jQuery
Platforms
iOS, Android, Google App Engine
Storage
JSON, Google Cloud Storage, Google Cloud Datastore, PostgreSQL
Education
2015 - 2017
Master's Degree in Computer Science
University of Warsaw - Warsaw, Poland
2012 - 2016
Bachelor's Degree in Mathematics
University of Warsaw - Warsaw, Poland
2012 - 2015
Bachelor's Degree in Computer Science
University of Warsaw - Warsaw, Poland
Collaboration That Works
How to Work with Toptal
Toptal matches you directly with global industry experts from our network in hours—not weeks or months.
1
Share your needs
Discuss your requirements and refine your scope in a call with a Toptal domain expert.
2
Choose your talent
Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.
3
Start your risk-free talent trial
Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.
Top talent is in high demand.
Start hiring