Ales Franek
Verified Expert in Engineering
Data Scientist and AI Developer
Ales is a data scientist with eight years of experience in machine learning, natural language processing, information retrieval, data analysis, and data visualization. He is also a strong Python programmer, writing readable, modular, and efficient code. Ales has often owned all parts of the development cycle, from product discovery to production deployment, focusing on impact, pragmatism, and finding creative solutions to business needs that deliver exceptional value.
Portfolio
Experience
Availability
Preferred Environment
PyCharm, DBeaver, Slack, iTerm2, Git, Zoom, Python
The most amazing...
...and impactful project I've worked on was a news recommender system for the most popular Czech web portal that's visited by millions of people every day.
Work Experience
Senior Data Scientist
Signal Media Limited, doing business as Signal AI
- Managed the end-to-end lifecycle and development of all NLP concepts, including discovery, research, data annotation, training, evaluation, serving, monitoring, reporting, and UX.
- Deployed new components to the production pipeline, which processed millions of documents per day.
- Optimized existing production pipeline services, thereby reducing their processing costs.
- Improved model evaluation in a model management platform containing thousands of live classifiers.
- Introduced a new way to classify documents, doubling the number of available categories.
- Played a key role in setting direction and OKRs as a member of a highly autonomous cross-functional team.
- Participated actively in hiring and onboarding new employees and chaired weekly research guild meetings.
Data Scientist
LexisNexis UK
- Significantly improved search relevance of the core legal research: established online and offline metrics; incorporated user engagement to the ranking algorithm; and enhanced query classification, recognition of legal phrases, and autocomplete.
- Educated the business on basic data science principles and promoted the data-driven decision-making culture.
- Oversaw data science initiatives across the UK division.
Applied Machine Learning Researcher
Seznam.cz
- Significantly improved performance and added new functionality to Seznam.cz, the most visited web portal and search engine in the Czech Republic.
- Improved components of the full-text web search engine and related services by using ML, NLP, neural networks, recommender systems, and anomaly detection.
- Built a framework for automated versioning, caching, parallelization, visualization, and reproducibility of data science experiments.
- Halved the error rate of body text extraction for crawled web pages, an essential part of web search.
- Increased the accuracy of learning-to-rank models by determining optimal discretization of continuous features for decision trees.
- Increased efficiency of the web crawler by predicting the times of likely future web page updates.
- Increased the click-through rate on news articles by developing a recommender for Seznam’s homepage, one of the most visited websites in the Czech Republic.
- Further improved the recommender by using article embeddings to mitigate the cold-start problem.
- Increased relevance of the autocomplete feature and designed an algorithm for semantic deduplication of the suggested queries.
- Developed a method for summarization of user product reviews for a comparison shopping service.
Computer Vision Researcher
Wikidi
- Designed and developed an end-to-end image retrieval system to detect brand logos in photos from social networks, capable of processing millions of images per day.
- Explored the potential for an ML-based video encoding algorithm.
- Contributed to a system that was able to find missing tech specs of products via internet searches.
Experience
News Recommender System for Seznam.cz
https://www.seznam.cz/Logo Recognition for Images From Social Networks
Machine Learning Meetups Prague
Skills
Languages
Python, SQL
Libraries/APIs
Matplotlib, Scikit-learn, Pandas, NumPy, SpaCy, XGBoost, Keras, MLlib, OpenCV
Paradigms
Data Science, Objectives & Key Results (OKRs), Extreme Programming
Other
Artificial Intelligence (AI), Machine Learning, Natural Language Processing (NLP), fastText, Tf-idf, Web Search, Website Ranking, Search, Information Retrieval, GPT, Generative Pre-trained Transformers (GPT), Software Development, Pattern Recognition, Data Analytics, Computer Science, Data Visualization, Decision Tree Regression, Decision Trees, Decision Tree Classification, Logistic Regression, Linear Regression, Neural Networks, SVMs, Support Vector Machines (SVM), Clustering, Recommendation Systems, Data Analysis, Image Processing, Predictive Analytics, Computer Vision, Game Theory, Combinatorial Optimization, Multi-agent Systems, Cryptography, Metaflow, Non-negative Matrix Factorization (NMF), t-SNE, Genetic Algorithms, Approximate Nearest Neighbors
Storage
Amazon S3 (AWS S3), DBeaver, PostgreSQL, Elasticsearch
Tools
PyCharm, Git, Kibana, Gensim, AWS Batch, MATLAB
Platforms
Docker, Amazon Web Services (AWS), Linux, Kubernetes
Education
Exchange Semester in Computer Science
National Taiwan University of Science and Technology - Taipei, Taiwan
Master's Degree in Artificial Intelligence
Czech Technical University in Prague - Prague, Czechia
Exchange Semester in Computer Science
University of Wisconsin–Madison - Madison, WI, USA
Bachelor's Degree in Cybernetics and Measurements
Czech Technical University in Prague - Prague, Czechia
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
Choose your talent
Start your risk-free talent trial
Top talent is in high demand.
Start hiring