Jan K.

Jan K.

Geneva, Switzerland
Hire Jan K. Hire Jan K. Hire Jan K.

About Me

Jan is a zealous machine learning engineer who possesses a fervor for software engineering, machine learning, leadership, and online education. He boasts a substantial magnitude of professional expertise in the utilization of computer vision, natural language processing, and time series analysis in both academic and business environments. In addition, Jan wholeheartedly devotes a significant portion of his leisure time to actively contributing to open-source software and engaging in the creation of educational content.

AI, ML & LLM

Pytorch Apache Airflow Machine Learning Machine Learning Operations (MLOps) Generative Pre-trained Transformers (GPT) BERT ChatGPT OpenAI GPT-4 API GPT Vertex AI Large Language Models (LLMs) OpenAI Containerization MLFlow LangChain Deep Learning Recurrent Neural Networks (RNNs)

Frontend

Backend

Database

DevOps

GitLab CI/CD Cloud Dataflow AWS Glue Terraform Google Kubernetes Engine (GKE) Kubernetes Docker Amazon Web Services (AWS) Google Cloud Platform (GCP) AWS Lambda Google Cloud Storage Amazon S3 (AWS S3) CI/CD Pipelines

QA & Testing

Workflow

Other

Python SAS Excel VBA Scikit Learn Numpy Keras SciPy Pandas Matplotlib Tensorflow AsyncIO Python AsyncIO OpenCV spacy Vim Text Editor Tensorboard Notion Amazon SageMaker Google Compute Engine (GCE) Inkscape Adobe Premiere Pro Seaborn Gensim StatsModels Scikit-image Data Science REST Scrum Agile software development Jupyter Notebook Amazon EC2 Elasticsearch Redis Redis Cache Probability Theory Mathematical Analysis Linear Algebra Statistics Portfolio Optimization Orchestration Shell Scripting Hugging Face Transformers Sphinx Natural Language Processing (NLP) Finance Computer Vision Artificial Intelligence Hugging Face Natural Language Understanding (NLU) Algorithms Back-end Development Language Models Pub/Sub Technical Leadership Leadership Retrieval-augmented Generation (RAG) Back-end Serverless Optimization Microeconomics Macroeconomics Mathematical Finance Quantitative Risk Analysis Numerical Methods JAX Time Series Analysis Product Consultant Measure Theory Econometrics Private Company Valuation Scrum Master Online Course Design Open Source Image Registration Data Versioning Google BigQuery Full Stack Development Apache Spark

Work history

Toptal
Machine Learning
2023 - Present (2 years)
Remote
The EPFL
Machine Learning Section Manager | Blue Brain Project
2023 - 2024 (1 year)
Remote
  • Managed a team of 4 experienced machine learning engineers

  • Designed a literature search system focused on semantic search, question answering, named entity recognition and entity linking built on top of recent large language models. The entire system was deployed at scale (Kubernetes, AWS).

The EPFL
Machine Learning Engineer | Blue Brain Project
2019 - 2023 (4 years)
Remote
  • Conceived and implemented a supervised algorithm for 2D brain slice image registration that became a part of internal workflows (https://github.com/BlueBrain/atlas-alignment)

  • Developed a knowledge extraction pipeline for scientific articles. Main functionalities include parsing, neural search and named entity recognition (https://github.com/BlueBrain/Search).

  • Direct involvement in other neuroscientific projects including neuron type classification with graph neural networks and morphology image synthesis with GANs

Nectar Financial
Data Scientist
2018 - 2018
Remote

Enhanced internal portfolio optimization algorithm with return forecasting using supervised learning techniques and I also added custom constraints and objective functions making the tool more flexible.

UBS
Quantitative Risk Analyst
2016 - 2017 (1 year)
Remote

Showcase

Mildlyoverfitted | Educational Videos
  • Spearheaded development of a YouTube channel, Mildlyoverfitted, showcasing educational content and resources

  • Channel features self-produced videos on machine learning, deep learning, and Python

  • Platform objectives include detailed explanations of various concepts and insights on implementing solutions from scratch

DeepDow | Portfolio Optimization with Deep Learning
  • Python package for portfolio optimization that combines market forecasting with LSTM and GARCH, and problem designing with convex optimization.

  • Constructs a layered pipeline where final layer allocates resources and preceding layers function as feature extractors.

  • Fully differentiable network that allows parameter optimization using gradient descent algorithms.

MLtype | Command Line Tool
  • Advanced command line tool designed to improve typing speed and accuracy for programmers

  • Enables programming language practice by using neural networks to generate text

  • Offers pre-trained networks or the option to train new networks from scratch

Atlas Alignment | Multimodal Registration and Alignment
  • Atlas Alignment is a toolbox made for multimodal image registration, providing both conventional and deep learning models.

  • The project was driven by the Blue Brain Project's endeavors in aligning atlas maps of mouse brain obtained from ISH gene expression and Nissl stains.

  • The work has been published on Frontiers Media and can be accessed via the provided link.

PyChubby | Automated Face-warping Tool
  • PyChubby is a Python package created for automating face warping processes.

  • It allows users to modify the facial expression and shape of individuals in images programmatically.

  • It utilizes computer vision techniques, particularly geometric transformations, to achieve its functions.

Education

Education
Master's Degree in Quantitative Finance
ETH Zurich
2015 - 2018 (3 years)
Education
Bachelor's Degree in Economics
Charles University
2011 - 2014 (3 years)