Tomás Mardones, Developer in Viña del Mar, Chile
Tomás is available for hire
Hire Tomás

Tomás Mardones

Verified Expert  in Engineering

Computer Vision Developer

Viña del Mar, Chile
Toptal Member Since
September 13, 2016

Tomás is an experienced engineer with great know-how in the fields of computer vision and machine learning. He's a researcher and an engineer who likes to build efficient and elegant software systems. Tomás is always eager to find better ways to write better code. In addition, he is an expert in C++, Python, and MATLAB.



Preferred Environment

Git, PyCharm, OS X

The most amazing...

...thing I've coded is a cross-platform iOS/Android 3D engine from scratch using OpenGL ES 2.0.

Work Experience

Site Reliability Software Engineer

2017 - PRESENT
  • Supported the YouTube infrastructure.
Technologies: Python, Google Cloud, C++

Temporal Full-Time Professor

2016 - 2017
Universidad Técnica Federico Santa María
  • Taught a "Pattern Recognition" postgraduate course for computer science students and "Introduction to Programming" for engineers.
  • Conducted a study regarding the position and the organization of the computer science department to gain insights about possible future directions of change.
  • Guided several pattern recognition student projects—ranging from image classification to data visualization.
Technologies: MATLAB, Python

Part-Time Professor

2015 - 2016
Universidad Técnica Federico Santa María
  • Taught "Introduction to Programming" course for engineers using Python.
  • Obtained some of the best evaluation scores from the students.
  • Every semester tried new ideas (new material, activities and videos) to engage the students with the course.
Technologies: Moodle, PyCharm, Python

Researcher (Ph.D Candidate)

2013 - 2016
Universidad Técnica Federico Santa María
  • Analyzed and corrected some widely-used assumptions related to the usage of Fisher Vectors in Image Retrieval, increasing the system precision by 5% using the same resources. (Published in ICPRAM 15).
  • Combined image descriptors using different sampling methods and attained a 12% precision increment in an Image Retrieval system. (Published in ICPRAM 15).
  • Proposed a new graph representation to combine many lightweight image representations achieving state of the art precision, while using 3 to 20 times less memory. (Published in CIARP '15 and JVCIR '16).
Technologies: MATLAB, Python, Dropbox, Bash, LaTeX, C

Co-Founder | Chief Technology Officer | Software Engineer

2012 - 2013
  • Designed and implemented a cross-platform (iOS, Android) C++ 3D graphic engine based on OpenGL ES 2.0 focused on the needs of augmented reality applications.
  • Implemented the dynamic load of 3D assets and textures to allow content to be downloaded and loaded on real-time, while taking into account the memory limitations of the device.
  • Incorporated DAE 3D models importation with morph animations.
  • Integrated Vuforia SDK (for augmented reality capabilities) with the 3D engine for iOS and Android.
Technologies: Bitbucket, Mercurial, Vuforia SDK, iOS, NDK, Android, Java, Objective-C, C++


2012 - 2012
Tom's Apps
  • Built a fast and responsive vignette viewer for a few websites (70,000 - 100,000 downloads).
  • The system built was 200% faster compared to competing applications, thanks to the use of simple cache techniques, using an custom parser instead of DOM parsers and managing the HTTP connections differently.
Technologies: Eclipse, Java, Android

Co-Founder | Chief Operating Officer | Software Engineer

2008 - 2012
  • Designed, implemented, and integrated the graphical back-end of a C++ framework to create flexible and extensible desktop Augmented Reality Applications.
  • Implemented a new graphical back-end using Flash 3D engines and integrated it with C++ cross-platform code via Alchemy.
  • Coordinated the assets creation with outsourced 3D artists and clients to fulfill their requirements, while keeping the technical constraints in check.
  • Wrote a Flash 10/11 API to freely create augmented reality applications using our technology.
  • Integrated FFmpeg for video playback on textures of 3D objects.
Technologies: Alchemy, Bitbucket, Subversion (SVN), OpenAL, Irrlicht, Away3D, Flare3D, OGRE, Flash, C++

Winner, First Place, Intel Global Challenge 2010 (Latin America)

First prize was awarded to the project that best attempted to revolutionize media via the extensive use of augmented reality.
2010 - 2016

PhD Degree in Computer Science

Universidad Técnica Federico Santa María - Valparaíso, Chile

2010 - 2014

Master's Degree in Computer Science

Universidad Técnica Federico Santa María - Valparaíso, Chile


NDK, OpenAL, D3.js


MATLAB, PyCharm, Vuforia SDK, Mercurial, Bitbucket, Moodle, LaTeX, Flash, Away3D, Subversion (SVN), Microsoft Visual Studio, Eclipse IDE, Git, Xcode


C++, Python, Objective-C, Java, C, Bash


Object-oriented Programming (OOP), Test-driven Development (TDD)


OGRE, Irrlicht, Alchemy


Google Cloud, MySQL


OS X, Eclipse, Dropbox, Flare3D, MacOS, Vuforia, Windows, Android, iOS


Machine Learning, Image Retrieval, Statistics, Mathematics, Computer Vision, Clustering, Data Analysis, Data Visualization, Information Retrieval

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.


Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.

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