Experienced machine learning engineer and software developer specializing in deep learning applications. Proficient in computer vision, medical device development, data mining and natural language processing. With expertise to take a project from the proof of concept all the way through to production, ensuring high-quality and effective solutions.
Data Analyst (contract), March - December 2022
- Performing data analysis of the company-wide risk-related data and automating report generation.
- Creating data pipelines to process and centralize data from multiple sources (Airflow).
- Building visual tools to enhance insight and facilitate early detection of threats (PowerBI, Dash).
Machine Learning Engineer (contract), June 2021 - February 2022
- Developed a semantic segmentation computer vision model for automatic evaluation of advanced driver-assistance systems (ADAS).
- Researched and evaluated the state of the art models (based on convolutional neural networks and vision transformers) and techniques to optimize semantic segmentation maps.
- Ported a Tensorflow based machine learning pipeline to OpenMMLab (PyTorch based) for the use on the local machines and on the cloud.
- Developed high-performance algorithms for post-processing of semantic segmentation maps.
- Assisted the acquisition of the driving dataset (images and LIDAR data) and built algorithmic and visual tools for optimizing the data selection.
Machine Learning Engineer (contract), April - May 2021
- Consulted on architecture improvements to the convolutional neural network-based deep learning model (PyTorch) and the pipeline for sustainable development and deployment in the production environment.
- Designed a REST API for client access to the model.
- Deployed the model as a service accessible through a REST API.
Machine Learning Engineer/Senior Data Scientist, May 2019 - December 2020
- Performed analysis and processing of large quantities of continuous patient data and conducted research on predicting severe outcomes in the intensive care setting.
- Developed a recurrent neural network based model for predicting post-operative complications.
- Designed the software architecture incorporating the machine learning model.
- Led and conducted the development in accordance with processes for the product to be certified under Medical Device Regulation.
- Integrated the product with the hospital’s infrastructure and deployed it
Data Engineer/Machine Learning Engineer, June 2018 - March 2019
- Led the effort to build a vast dataset of medical mammography images in collaboration with multiple healthcare sites in Europe.
- Anonymized and processed hundreds of thousands of images for use in a machine learning project.
- Developed the backend for an online platform for X-ray image annotation.
- Participated in development of a convolutional neural network based deep learning model for detecting and classifying abnormalities in mammography images.
/Machine Learning Engineer/Research Programmer, November 2013 - May 2018
- Conducted research on computational representation learning and its application to electronic medical records.
- Developed supervised models using recurrent neural networks, noisy medical data to predict multi-label targets.
- Designed and developed a web application for visualization of medical history employing semantic embedding for medical concepts.
- Compared efficacy of different data formats and common predictive models for medical outcome prediction.
- Implemented algorithms for statistical modeling in Python and C optimizing for speed and API simplicity.
- Configured and maintained a GNU/Linux based computation server.
Software Developer, September 2012 - November 2013
- Developed and maintained several web applications for human genetic research data management and collaboration.
- Designed and developed mobile applications for iOS and Android including a computer vision program for genetic sample management.
- Configured and maintained a GNU/Linux based web and database servers.
Software Developer/Research Assistant, September 2010 - September 2012
- Developed large structural biology oriented databases and web applications for research collaboration and communicating results to the public.
- Configured and administered a network of highly utilized GNU/Linux based servers.
- Expanded the web-based laboratory information management system, by providing unified data storage, improving data visualization and sharing capabilities in structural genomics laboratories.
- Developed a high-throughput pattern recognition system, which automatically scans images in search of protein crystals.