cv

Contact Information

Name Antoni Kingston
Professional Title Computer Science student
Email kingstonantoni2003@gmail.com
Location Warsaw,

Professional Summary

Computer Science student at Warsaw University of Technology and intern at CCAI. Interested in model science and interpretability based on representation geometry in LLMs.

Experience

  • 2025 - 2025

    Warsaw, Poland

    Data Warehouse Administrator (Internship)
    Accenture
    • Analyzed long SQL queries to deal with reported inconsistencies in telecommunications DWH
    • Developed a Django web app for custom ticket management based on the system’s API
    • Got to know the inner workings of DWH operations, especially of the ELT process

Projects

  • gipsDA R library

    An R package for discriminant analysis classification using covariance matrices with permutation symmetries

    • A fully functional R library available on CRAN, developed with best programming practices
    • Enhances the functionality of gips by allowing estimation of permutation symmetry shared by multiple matrices
    • Offers functions to train Linear and Quadratic Discriminant Analysis classifiers utilizing gips methodology
  • Tweet Sentiment Predictor

    Binary sentiment classification on 100k Sentiment140 tweets via NLP feature engineering and ensemble ML

    • Built a three-stage Jupyter pipeline (feature engineering, selection, modeling) on a balanced 100k-tweet corpus with reproducible 5-fold cross-validation
    • Engineered lexical, readability, temporal, and embedding features using VADER, TextBlob, Word2Vec, Sentence-BERT (all-MiniLM-L6-v2), and spaCy; selected predictors with L1 logistic regression and feature_engine
    • Tuned and compared XGBoost, LightGBM, voting/stacking ensembles, and a Keras MLP with Optuna, achieving ~81% cross-validated accuracy on the extended feature set
  • A Simple GPT model

    Generative language model training

    • Implemented a causal Transformer language model in PyTorch (attention blocks + RMSNorm)
    • Built a custom BPE tokenizer and corpus-to-encoding pipeline; trained on Adam Mickiewicz works and generated text via generate(…)
    • {“In progress”=>”scaling training to the Polish Parliamentary Corpus (PPC)”}

Education

  • 2026 - Present

    Warsaw, Poland

    Master's Studies
    MiNI WUT
    Computer Science and Information Systems
    • Methods of Artificial Intelligence Specialization
  • 2022 - 2026

    Warsaw, Poland

    Bachelor's Studies
    MiNI WUT
    Data Science and Engineering
    • Adapting the gips Library for Classification Problem Utilizing Discriminant Analysis (Bachelor’s thesis)

Skills

Technologies: Python (numpy, pandas, sklearn, pytorch, matplotlib), R, SQL, Bash, basic knowledge of C++, Java and HTML/CSS
Mathematics: Linear Algebra, Probability Theory, Statistics, Stochastic Processes, Machine Learning, Financial Mathematics, Convex Optimization, Algorithms and Data Structures

Languages

Polish : Native
English : C2 (CAE Certificate)
German : B1

Certificates

  • Cambridge English (CAE) - Cambridge Assessment English