Felicia Segui

Ongoing: App for crafts and DIY-projects

Currently, me and another software developer are creating a social media app for crafts and DIY-projects. The idea is to have a platform on which the users can share their DIY-projects, as well as find inspiration, guides and patterns for new projects. Both me and my partner are into fiber arts and we want to simplify documenting the creative process. The application is written in dart/Flutter, using a NoSQL database.

• Code

Real-time unsupervised log event anomaly detection in public transportation

In my master's thesis, I developed a novel method for detecting anomalies in public transportation bus event log data in real-time, without the need for a labeled dataset. By using a combination of an autoencoder, PCA model, and clustering algorithm, this research successfully identifies anomalies within event sequences. The best-performing model, combining the autoencoder and clustering algorithm, achieved a remarkable F1-score of 0.79. This work contributes to the field of real-time anomaly detection, using combined Machine Learning Models, to enhance public transportation system reliability and security.

• Report

cliMind

In the beginning of the autumn 2021 me and two friends started a Machine Learning project for predicting stock market fluctuations, based on weather forecasts. This idea came to us from reading articles about the possible correlation between sunny weather and a rising stock market. We thought, why not test the hypothesis? We explored many different ML models and decided to dig deeper into two of them; an Azure AutoML Timeseries model and a Tensorflow Neural Network model.

• Code

Computer Graphics project: Game

As a final project in a computer graphics course me and another student made an OpenGL game together. Here I got the chance to use my math skills and at the same time expand my software knowledge. In addition to consolidate my C++ writing, I learned how to write pixel shaders, how to navigate between different coordinate systems and the many aspects of creating a well working game.

• Code

Ongoing: Audio and ML

I am currently experimenting with different audio datasets in ML-algorithms, as it is an area that really interests me. I aspire to soon dig deeper into music genre classification so feel free to check out what I am working on right now!

• Code

Database modeling project

Together with two other students I created an SQL database for a fictional large scale bakery to keep track of its production and deliveries. The project contained three major parts: a planning part in which we created an ER-model, an SQL part and a part for implementing a REST server.

• Code

LSTM Networks: NER-Recognition

By using recurrent neural networks two models were created and trained to recognize named entities in texts. The first model was a simple recurrent neural network and the second model an LSTM network. From each model two new improved models were created by adding dropout and bidirectional layers.

• Code

Convolutional Networks: Image-Recognition

In this project I used Convolutional Neural networks for determine what kind of flower an image showed. Four different models were built and compared with each other with plots of the train and validation accuracy over epochs, confusion matrices and accuracy scores. Both augmented and unmodified images were used, as well as a pretrained convolutional neural base.

• Code

Decision trees: Implementation of the ID3 classifier

In the course Applied Machine Learning we implemented an ID3 algorithm to generate decision trees, by using recursive methods. The algorithm was then trained and used for classifying handwritten digits. Before implementing the ID3 algorithm we used the already existing Scitkit-learn DecisionTreeClassifier for classifying the same dataset and in the end of the assignment the different f1-scores were compared.

• Code

Scitkit-learn: Extracting Syntactic Groups using machine-learning techniques

In this project we created a program that extracted syntactic groups from a text. We compared a simple non-ML model with a supervised ML-model and further improved the ML-model by adding two dynamic paramaters to the feature vector. From the resulting dataset we extracted a smaller set and implemented a method to find the entities on Wikipedia.

• Code
• Report

Python: Analyzing 911 Calls and Finance

Before starting with the Machine Learning part, the course "Python for Data Science and Machine Learning Bootcamp" provided two final projects, which involved data analysis with Pandas and NumPy, as well as visualization tasks where we were encuraged to use Plotly, Seaborn and Cufflinks. Especially in the finance project, I realized the importance of having visualization tool skills, as many interesting patterns first became visible after the data underwent different data extractions and then shown with carefully chosen graph-types.

• Code 911 Calls project
• Code Finance project

Matlab: Finite element method: Netflix microchip

In the course Finite Elementh Method me and a friend investigated how a microchip in Netflix's circuits, was affected by the decreased streaming quality which was implemented when the streaming consumption increased during the COVID-19 pandemy. The analytical part was performed in Matlab and the chip was modeled with the finite element method. The resulting plots showed the temperature distribution with and without decreased streaming quality, as well as the displacement fields due to thermal expansion.

• Code
• Report

Java: Sudoku solver

As a final project in one of the university's Java programming courses me and a friend built a sudoku solver by implementing a recursion- and backtracking algorithm. In addition to the recursion and backtracking we improved our Java Swing knowledge and learned how to effectively write code together as a group.

• Code

• Github
• LinkedIn
• -E-mail

Ongoing: Senior Software Engineer & Data Analyst at ABCyber

In December 2024, I joined ABCyber as a Senior Software Engineer and Data Analyst. In this role, I work as a full-stack developer, focusing on the development and enhancement of the company’s web applications. My responsibilities include improving existing features, creating new functionality, and integrating data analytics to streamline daily maintenance and reporting processes.

Technologies/Frameworks/Languages: Vim, Php, HTML, CSS, JavaScript, Vue, Laravel, Docker, Kubernetes

IT Consultant at Netcompany A/S

In September 2022 I started working as IT Consultant at Netcompany in Copenhagen. During my time at Netcompany, I worked in two projects: as backend developer at the Danish taxation authorities and with data management for financial and HR reporting.

In both projects, the teams were small and I got experience from all steps in the software lifecycle, from analysis and design, to implementation, testing, deployment and maintenance.

During the last part of my employment, I got promoted to Team Leader for the global internal data management team.

Technologies/Frameworks/Languages: PostgreSQL, Java, RabbitMQ, Docker, Kubernetes, Openshift, Jenkins, IBM MQ, Splunk, SQL, Scala, Gradle, Groovy, Grafana, Python, Scrum (Safe Framework), Gatling, Git, DAX, PowerBI, Microsoft SSMS, OLAP Cubes

Part-time Systems Developer at Gaia System AB

In September 2021 I started working part-time at Gaia Systems AB, in parallell to my studies. It is the same company as I did an internship at in spring 2021 and I continued on the Machine Learning project that I worked on at the internship, together with other ML and AI oriented projects. I worked here until January 2022 when I replaced my studying and part-time working with Master's thesis writing.

My main tasks during my time at Gaia were building, training and evaluating ML models in Azure Machine Learning Studio, extracting and transforming big datasets using PySpark in Azure Synapse, as well as link and automate the different parts. Scroll down to read more about the project I was working on and my internship experience.

Technologies/Frameworks/Languages: Azure ML Studio, Azure Synapse, PySpark, Pandas, Seaborn, PyTorch, Keras, Tensorflow, SQL, Docker, Git

R&D Summer Internship at Ericsson

The summer 2021 I got the opportunity to work at the research department with high precision satellite positioning, using cellular network assistance data. Here I worked on the client side and it was very stimulating to combine programming with my physics and math knowledge at a whole new level.

The code was written in C in Ubuntu which I before beginning thought would be a challenge, but after approximately a week I felt confident with the code syntax. Except for thorougly learning C as well as how to work with precision satellite positioning, I developed the ability to quickly get a good overview of large code files. This I had to develop because I needed to search for code with a certain functionality in large repositories. Sometimes this was really challenging due to differences in documentation and definitions between the repositories, but in the end super rewarding and a skill I will have use for in the future.

Technologies/Frameworks/Languages: C, C++, Ubuntu, Raspberry Pi, GNSS Receiver, Git

ML Internship at Gaia Systems

In spring 2021 I went on an internship at Gaia Systems, a company that is specialized in Data analytics, IoT and AI. Gaia Public Transport is a passenger information system and currently the main project at Gaia. The system is a cloud-based real-time system that facilitates travels for passengers, drivers and operators and also the project that I have mainly been a part of this internship.

My work tasks during this internship were machine learning and AI oriented. I mainly worked with the previously mentioned project Gaia Public Transport, but also with a smaller computer vision project that aims to classify plant weed, plant diseases and plant vermin. The prior goal for me in this internship was to develop an Azure ML-model that predicts bus stop arrival times for Gaia Public Transport. Since the project I worked on was in a research state, everything except the goal was very open and the type of model, features, limits and parameters were specified as time went. Because of this my tasks were very varied, within the ML area. I had tasks such as pre-research, data extraction and transformation, model building and evaluation of the model, among others.

In the pure technical area the most valuable experience for me was working cloud based in Azure, with the services that were included. Azure ML Studio and Azure Synapse were the platforms that I mainly worked on, which was super interesting as I for the first time got the possibility to apply my knowledge from University at real life cases. Apart from purely technical aspects I have learnt a lot about how it is to work with software development in bigger groups, as well as alone remotely during the Covid-19 pandemy. As the proceeding of my project was in some ways dependent on other colleagues, I got both insights in software areas other than machine learning, since it has been crucial that I understand what is going on when my colleagues works with their parts, as well as a better understanding for the whole process on a bigger level.

Down below is a link to the final presentation I made at Lund University, as well as a link to an article that was written about my internship.

Technologies/Frameworks/Languages: Azure ML Studio, Azure Synapse, PySpark, Pandas, Seaborn, PyTorch, Keras, Tensorflow, SQL, Docker, Git

• Presentation
• Article

Hi and welcome to my portfolio :)

This website is usually my software development portfolio, but I figured I can use it to provide some valuable info for my application to volunteering at waking life 2025. I will present some of my prior experiences relevant for the volunteering, but please have a look at the application form, to see my team/role-specific motivations.

Experiences

• Portfolio

Text

This is bold and this is strong. This is italic and this is emphasized. This is ^superscript text and this is _subscript text. This is underlined and this is code: for (;;) { ... }. Finally, this is a link.

Heading Level 2

Heading Level 3

Heading Level 4

Heading Level 5

Heading Level 6

Blockquote

Fringilla nisl. Donec accumsan interdum nisi, quis tincidunt felis sagittis eget tempus euismod. Vestibulum ante ipsum primis in faucibus vestibulum. Blandit adipiscing eu felis iaculis volutpat ac adipiscing accumsan faucibus. Vestibulum ante ipsum primis in faucibus lorem ipsum dolor sit amet nullam adipiscing eu felis.

Preformatted

i = 0;

while (!deck.isInOrder()) {
    print 'Iteration ' + i;
    deck.shuffle();
    i++;
}

print 'It took ' + i + ' iterations to sort the deck.';

Lists

Unordered

• Dolor pulvinar etiam.
• Sagittis adipiscing.
• Felis enim feugiat.

Alternate

• Dolor pulvinar etiam.
• Sagittis adipiscing.
• Felis enim feugiat.

Ordered

1. Dolor pulvinar etiam.
2. Etiam vel felis viverra.
3. Felis enim feugiat.
4. Dolor pulvinar etiam.
5. Etiam vel felis lorem.
6. Felis enim et feugiat.

Icons

• Twitter
• Facebook
• Instagram
• Github

Actions

• Default
• Default

• Default
• Default

Table

Default

Alternate

Buttons

• Primary
• Default

• Default
• Small

• Icon
• Icon

• Disabled
• Disabled

Form