Personal details

Vasileios - Remote

Vasileios

Timezone: Athens (UTC+3)

Summary

Data Scientist with a clear, logical mind and a practical approach in problem solving. Open-minded, ambitious and eager to learn person, with genuine interest in Data Science & Deep Learning areas. Acquired a wealth of knowledge and skills through work experience in various creative industries and academic roles.
Expertise in data pre-processing methods and training high quality and robust models to decipher meaningful patterns from data, in the most effective way possible, taking client relationship very seriously.

Work Experience

Data Scientist
Deutsche Telekom | Jun 2021 - Present
Machine Learning
Python 3
Data Science
Deep Learning
TensorFlow
Data Modelling
PyTorch
- Working on personalised recommendation engines for MagentaTV DE. - Serve high quality personalised recommendations to the users using state-of-the-art approaches like xDeepFM - Data conditioning and feature engineering
Data Scientist
Propulsion Analytics | Jan 2019 - Present
Keras
The main responsibilites of my current role as a data scientist are: - Extract insights from timeseries data through careful analysis. - Perform data cleansing and transformation. - Train, validate and explain machine learning model performance. - Deploy machine learning models. - Communicate the results to the stakeholders.

Personal Projects

Malicious Network Traffic DetectionIconOpenNewWindows
2018
Pandas
- Performed EDA on the dataset. - Cleaned the erroneous observations and performed the necessary statistical transformations. - Anlayzed the data and extracted insights that I later communicated to the stakeholders. - Train and evaluated the appropriate ML models. - Wrote the deployment code.
Physical Distaster Classification on Twitter text data
2019
NumPy
Pandas
Keras
During the development of that project I assisted in the following tasks: - Data acuisition. - Text cleaning and transformation (Word2Vec/Doc2Vec). - Text analysis and predictive modelling (CNN & traditional ML models).