Stanisław Wilczyński

React developer delivering high-value experiences across the Microsoft 365 ecosystem, used by millions of users daily. Responsible for:

• Leading the development of innovative products in AI space (e.g People Agent), providing architectural direction for a cross-geo group of ~40 developers
• Driving technical execution across feature crews and bridging product, design, and multiple engineering teams
• Designing and implementing user-facing products (e.g. People App) with a strong focus on accessibility, performance and maintainability
• Being organization-wide expert in building and improving testing and dependency management tooling (both open-source and internal) across large monorepos, enhancing developer productivity at scale, contributing to multiple RFCs in both areas
• Onboarding new team members, leading knowledge-sharing sessions, and mentoring developers across the organization

Working as Java developer on maintaining and improving one of the top European logistics SaaS platforms. Projects I took part in included:

• Implementation of refined and redesigned UI for application main components
• Exhaustive improvement and refactoring of aliasing mechanism for tailored user experience
• Self-directed work improvements:
  • unification and enhancement of local deployment tools
  • integration of code review platform with our workflow
  • introduction and application of User Story Mapping and Pair Programming in the team

Helping other students prepare for the exams

Final grade: 5.0/5.0

Thesis: Microarray data analysis in prediction of breast cancer metastasis. Here you can find code and the thesis itself.

Final grade: 5.0/5.0

Thesis: Reduction of dimensionality by sparse subspace clustering

On my github you can find a list of all courses I took during my studies.

One of the biggest statistical conferences in Europe. I gave a talk on Multiple latent components clustering.

Continued working on Multiple Latent Components Clustering (MLCC). The collaboration with University of Angers aimed at extending the method to work with data from zero inflated negative binomial distribution and proving mathematical properties of MLCC and PEnalized SEmi-integrated Likelihood (PESEL) method. Furthermore, we use MLCC to extract significant features/information from huge breast cancer data set (1200 observations, 60000 variables).

As one of just 8 people from across Europe, I was selected to present results of my research on Young Statisticians Session. I gave a talk on Reduction of dimensionality by sparse subspace clustering.

In this project I started working on Multiple Latent Components Clustering (MLCC). I was responsible for formalizing its mathematical basis, maintaining and extending R package Varclust that implements the method. Moreover, I performed throughout simulation study to justify that MLCC is competitive with state-of-the-art variable clustering methods.