Welcome to my portfolio :) Here, you can find info on my recent work in NLP, ML, computer vision and data engineering, and the links to their GitHub repos.
Independent project (with G. Samo, S. Fuzhen)
Health-related content in transformer-based language models
A research project in which we leveraged our theoretical linguistics and NLP skills to investigate the accessibility of health communication during the COVID-19 outbreak.
Tools: Python, Hugging Face Transformers, Hugging Face Datasets, MatPlotLib, R.
Omdena Local Chapter - Lublin
Conversational AI Chatbot for People Affected by High Inflation and Increased Cost of Living
Community project in which I am collaborating that aims at the creation of a conversational chatbot that will be able to provide financial advice to those affected by the current economic crisis.
OpenSouce (launches on February 2023)
A community project that utilises data engineering and NLP/NLU techniques to create an Open Souce corpus of spoken interactions in the Romance languages. The project aims at solving the problem of data scarcity and data reliability in linguistic investigations.
Omdena Local Chapter - Marseille
Deploying an Accurate Classifier to Stop Online Violence Against Children using NLP
Community project in which I am collaborating that aims at stopping on-line violence against children by deploying an accurate classifier to identify grooming behaviour in online chats with children.
Facebook Marketplace Recommendation Ranking System
Specialist project on which I am working as a part of my training in AI and Data at AiCore.
@ Currently training a model that works with images and text to make recommendations to potential buyers based on their demographic information;
@ Will host the app on EKS to orchestrate the containers that maintain/retrain the model.
Multinational Retail Data Centralisation
Data collection project on which I have just started working as a part of my training in AI and Data at AiCore.
Web Scraping Pipeline
Web scraping project on which I worked as a part of my training in AI and Data at AiCore.
@ Developed a module that scrapes data to upsert information to a database;
@ Pushed an image of the containerised application in a CI/CD pipeline that performs unit testing and deploys the image to an EC2 instance.
Computer Vision Application
Computer vision application on which I worked as a part of my training in AI and Data at AiCore.
@ Trained a computer vision model and used Tensorflow to detect whether Rock, Paper or Scissors is shown to the camera in real-time and with a high accuracy.
@ Used the OpenCV library to access the webcam and play Rock Paper Scissors with the computer using the image from the camera.
Use a Pre-Trained Image Classifier to Identify Dog Breeds
Project in which we tested our Python skills by determining:
@ which Pytorch image classification algorithm among AlexNet, VGG, and ResNet performs best on classifying images as dogs or not dogs.
@ how well this best classification algorithm works on correctly identifying a dog's breed.
@ how long each algorithm takes to solve the classification problem (runtime).