Kristian is available for hire
Hire KristianKristian Kozak
Verified Expert in Engineering
Software Developer
Location
Brno, South Moravian Region, Czech Republic
Toptal Member Since
January 21, 2022
Kristian is a senior developer focused on data engineering and web development. He has been building software for almost a decade, has leadership experience from a five-person research and development team, and has two published papers under his belt. Kristian specializes in working with data in Python and R and developing web services in Ruby on Rails.
Portfolio
EnliteAI
Python, NumPy, PyTorch, Machine Learning, Concurrent Programming...
Wandera (acquired by Jamf)
R, Python, Data Analysis, Data Visualization, Machine Learning, Docker, Jira...
Wandera (acquired by Jamf)
R, Python, Tidyverse, SQL, Data Analysis, Machine Learning, Data Visualization...
Experience
Availability
Part-time
Preferred Environment
PyCharm, RStudio, Git, Slack, Jira, Basecamp, MacOS, Sublime Text
The most amazing...
...project I've worked on was designing and developing an applied reinforcement learning framework focused on real-world problems from scratch.
Work Experience
Senior Machine Learning Engineer
2020 - PRESENT
EnliteAI
- Participated in developing the open-source reinforcement learning framework MazeRL from scratch, designing and building complex components such as experience module or agent deployment, with a lot of focus on efficiency and parallelism.
- Co-created a web application presenting interactive AI-assisted electrical grid control scenarios, showcasing its capabilities to prospective clients.
- Used the developed RL framework in client projects and challenges such as the "Learning to Run a Power Network" (L2RPN) organized by RTE and TenneT, French and German grid operators.
Technologies: Python, NumPy, PyTorch, Machine Learning, Concurrent Programming, Software Engineering, Data Science, Amazon Web Services (AWS), Pandas, Jupyter Notebook, Artificial Intelligence (AI), Unit Testing, Git, Data Engineering, Data Analysis, Data Visualization, Data Analytics, Deep Learning, Reinforcement Learning, JavaScript, HTML, CSS, R, Data Wrangling, System Architecture, Architecture, Graphs, Flask, Web Development, Back-end
Team Lead, Research and Development Team
2018 - 2019
Wandera (acquired by Jamf)
- Led a team of five developers and researchers, prioritizing work and setting short- to mid-term goals concerning improving the research output of the team and the security capabilities of Wandera's threat defense platform.
- Prioritized and facilitated internal tools development that helped the team collect and process more data, freeing up more time for research and leading to better-published results and security features integrated into the Wandera platform.
- Coordinated responses to security incidents related to Wandera's platform and protected devices such as the iMessage vulnerability in the summer of 2019, mitigating the impact on clients and Wandera's business.
Technologies: R, Python, Data Analysis, Data Visualization, Machine Learning, Docker, Jira, Agile Sprints, Leadership, Feature Prioritization, Software Engineering, Data Processing, Data Science, SQL, Redash, Shell Scripting, Amazon Web Services (AWS), VPN, Threat Intelligence, Business Intelligence (BI), Data Modeling, Mobile Security, Amazon S3 (AWS S3), Pandas, Jupyter Notebook, Web Scraping, Data Scraping, Scraping, Artificial Intelligence (AI), Unit Testing, Git, APIs, Data Engineering, Data Analytics, JSON, Data Wrangling, System Architecture, Architecture, Microsoft Excel, Charts, Graphs, Databases, Flask, Web Development, MongoDB, Back-end, Integration
Data Engineer, Research and Development Team
2018 - 2019
Wandera (acquired by Jamf)
- Led the intelligence database effort in designing, building, and maintaining a database that was crucial for bringing the automation in the team to the next level, resulting in more time for research.
- Created an internal app analysis service, helping researchers automate parts of application analysis, ensuring they have all required information at hand when examining a mobile app.
- Worked with data collected from Wandera's mobile threat defense platform protecting companies like Santander or MasterCard, drawing insights and building PoC machine learning classifiers for improving Wandera's detection.
- Built data pipelines and automation around the intelligence database, including processing 1000+ Google sheets from past investigations or integrating data from third-party services.
- Presented research insights and handed off prototypes to product teams for implementation.
Technologies: R, Python, Tidyverse, SQL, Data Analysis, Machine Learning, Data Visualization, Scikit-learn, Docker, PostgreSQL, Software Engineering, Data Processing, Data Science, Amazon Web Services (AWS), Redash, Shell Scripting, VPN, Threat Intelligence, Mobile Security, Amazon S3 (AWS S3), Pandas, Jupyter Notebook, Web Scraping, Data Scraping, Scraping, Artificial Intelligence (AI), Unit Testing, Git, APIs, Data Engineering, Data Analytics, JSON, Data Wrangling, System Architecture, Architecture, Microsoft Excel, Charts, Graphs, Databases, Business Intelligence (BI), Flask, Web Development, MongoDB, Back-end, Integration
iOS Developer
2015 - 2018
Hlídacky.cz, s.r.o
- Developed the iOS app as a single developer, from the first design to launch on the App Store and further development.
- Implemented a complete feature set from browsing available sitters and their profiles, through real-time chat and job ads, to in-app payments and membership purchases, bringing new customers to the platform.
- Coordinated web API development with the Rails developer team of the core web application.
Technologies: iOS, Objective-C, Swift, Xcode, Ruby on Rails (RoR), Software Engineering, Unit Testing, Git, APIs, Heroku, JSON, Software as a Service (SaaS), Architecture, System Architecture
Experience
Paperback
http://paperbackhq.comA small SaaS business that helps teams save their time and attention.
I was the sole developer behind the application, from the front end through background jobs to payments and subscription processing.
I was the sole developer behind the application, from the front end through background jobs to payments and subscription processing.
MazeRL
https://github.com/enlite-ai/mazeMazeRL is an applied reinforcement framework aimed at real-world projects.
As part of the amazing team at EnliteAI, I designed and implemented individual components of the framework from scratch, such as experience module or agent deployment, and improved upon components created by others.
As part of the amazing team at EnliteAI, I designed and implemented individual components of the framework from scratch, such as experience module or agent deployment, and improved upon components created by others.
eCommerce Pot
https://ecommercepot.comA small Shopify agency focusing on custom development, marketing services, and Shopify app development. I am currently co-running the agency with my partner, handling Shopify app development and the more complex theme development tasks.
CopyCrab: Product Descriptions Writer
https://ecommercepot.com/apps/copy-crab-product-description-generatorCopyCrab is a Shopify app helping merchants write product descriptions with the help of OpenAI's GPT API. Merchants provide a few product highlights and keywords or brief descriptions, and CopyCrab turns these into a compelling, SEO-friendly product descriptions. The merchant can then edit the description to their liking and publish it on their store with a single click. I acted as the sole developer behind this application.
Data Job List
http://www.datajoblist.comA simple web app that aggregates jobs in data science, data engineering, AI, and related fields. The open positions are scraped and fetched from multiple sources around the web and combined into a clean, readable interface, and posted to Reddit and Twitter.
Issued for Abuse: Measuring the Underground Trade in Code-signing Certificates
https://weis2018.econinfosec.org/wp-content/uploads/sites/5/2018/06/WEIS_2018_paper_14.pdfA measurement study of underground trade in Windows code-signing certificates.
As the leading author, Kristian extracted and analyzed code-signing certificates from over 10,000 malware samples and scraped an underground marketplace with code-signing certificates, monitoring its sales and revenue over 100 days in five-minute intervals. The results have shown a stable demand and underground supply of abused code-signing certificates.
I started this study at the University of Maryland, working on it beyond the internship and then presenting it at the WEIS 2018 conference.
As the leading author, Kristian extracted and analyzed code-signing certificates from over 10,000 malware samples and scraped an underground marketplace with code-signing certificates, monitoring its sales and revenue over 100 days in five-minute intervals. The results have shown a stable demand and underground supply of abused code-signing certificates.
I started this study at the University of Maryland, working on it beyond the internship and then presenting it at the WEIS 2018 conference.
The Broken Shield: Measuring Revocation Effectiveness in the Windows Code-signing PKI
https://www.usenix.org/conference/usenixsecurity18/presentation/kimA measurement study of the effectiveness of code-signing certificate revocation was presented at USENIX Security 2018, a top-tier security conference.
During an internship at Maryland Cybersecurity Center (at the University of Maryland), I contributed to this study by helping to process and analyze a dataset of over 100,000 certificates used to sign Windows malware, detecting and classifying problems in the code-signing ecosystem, such as failed abuse detection, late revocation, or erroneous revocation date setting.
During an internship at Maryland Cybersecurity Center (at the University of Maryland), I contributed to this study by helping to process and analyze a dataset of over 100,000 certificates used to sign Windows malware, detecting and classifying problems in the code-signing ecosystem, such as failed abuse detection, late revocation, or erroneous revocation date setting.
Skills
Languages
Python, R, Ruby, SQL, JavaScript, CSS, HTML, Swift, Objective-C
Frameworks
Ruby on Rails (RoR), Stimulus, Tailwind CSS, Flask
Libraries/APIs
NumPy, Tidyverse, Ggplot2, Scikit-learn, Pandas, PyTorch, Sidekiq, Stripe API, Twitter API, Reddit API, Stripe, Shopify API
Paradigms
Data Science, Concurrent Programming, Business Intelligence (BI), Unit Testing
Other
Data Analysis, Data Visualization, English, Software Engineering, Data Processing, Threat Intelligence, Mobile Security, Data Analytics, Data Wrangling, Full-stack, System Architecture, Architecture, Charts, Graphs, Web Development, Back-end, Responsive UI, Concurrency, Secure Coding, Machine Learning, Web Security, Agile Sprints, Leadership, Feature Prioritization, Web Scraping, Data Scraping, Scraping, Artificial Intelligence (AI), APIs, Data Engineering, Software as a Service (SaaS), Hotwire, Reinforcement Learning, Shell Scripting, Data Modeling, Deep Learning, Integration, Shopify Theme
Tools
PyCharm, Jira, Git, Slack, Basecamp, Redash, Turbo, LaTeX, Xcode, VPN, Sublime Text, Microsoft Excel, RSpec
Platforms
RStudio, Jupyter Notebook, iOS, MacOS, Docker, Amazon Web Services (AWS), Heroku, Shopify
Storage
JSON, Databases, PostgreSQL, Amazon S3 (AWS S3), MongoDB
Education
2016 - 2019
Master's Degree in Computer Science
Masaryk University - Brno, Czech Republic
2017 - 2017
Research Internship in Computer Science and Data-driven Security
University of Maryland - College Park, MD, USA
2017 - 2017
Erasmus Student Exchange in Computer Science
Johannes Kepler Universität - Linz, Austria
2013 - 2016
Bachelor's Degree in Computer Science
Masaryk University - Brno, Czech Republic
2015 - 2015
Erasmus Student Exchange in Computer Science
Technische Universität Wien - Vienna, Austria
Certifications
JANUARY 2013 - PRESENT
Cambridge ESOL CAE – Grade A (Level C2)
University of Cambridge
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