Java-udvikler

where I solve software tasks for clients, both independently and as part of teams.

it-minds.dk +45 71 99 28 88 - Company IT Minds Period 09/2021 - Now:

 Project: Maintaining, servicing and updating
 Customer: University College Nordjylland
 Project Description: In the process of the customer phasing out teams and solutions to third-party solutions, I worked on learning their architecture, systems and solutions in a short amount of time. This was done with the goal of being able to maintain and service them during Technologies: the process. This included understanding the system architecture and use cases for ASP.NET, SharePoint,different solutions, as well as reading and understanding the code behind. The different solutions were written in C#/.Net and made use of Sharepoint, Microsoft Graph and / .NET Power Automate Processes.

My Function & Responsibility During the project: I worked collaborative with the development team to learn about each system and application that they maintain as well as their place in the overall structure of systems. In relation to this, I was responsible for reading, updating and testing the backend code for different solutions, with the goal of being able to maintain, update or write new code. Alongside doing this, I also wrote documentation to help others more easily maintain their solutions.

Project: Backend Python application
Customer: Vestas
Project Description: In a team of two, I built a backend application in Python, consisting of an API, Master node, Worker node and Predictor node. The application used custom configuration files to automatically build and train various types of Machine Learning models to predict future time-series data, keeping them all up-to-date based on time intervals. Custom configuration files to control various training parameters were added through the API. The Master node controlled the scheduling of training and updating the model. The worker node trained the specific model on the data, and the predictor node kept the future predictions up-to-date, storing the results in an influx database. The solution was written in Python and made use of Tensorflow, RabbitMQ and used an InfluxDB and SQlite database.

My Function & Responsibility: As part of the project, I provided insight in Machine Learning practices and built the Worker node responsible for training the models to predict time-series data. I also implemented the configuration file system used throughout the application as well as the methods to interact with an influx database. Moreover, I helped with various methods in the Master and Predictor node.

Title Computer Science During my studies and spare time, I worked on several projects to Student gain experience in the field of software development.
Company Non-commercial projects Period 09/2016 - 06/2021 Project Machine Learning project Customer: Self-employed Technologies: Python, TensorFlow Project Customer: Self-employed Technologies: PostgreSQL, C# / .NET Project Customer: Self-employed Technologies: Java + J2EE Project Description In colaboration with five other developers, I worked on a Machine Learning project wherein I researched, developed and implemented two Machine Learning models from Google DeepMind, Deep Q-learning and A3C. All which are capable of learning to play Atari2600 video games at a human level or above. In addition to this, the project also revolved around designing and implementing a third model which used the best of the two other Machine Learning models in an attempt to reach better results.
Function & Responsibility During the project two others and I were responsible for researching and implementing both a Deep-Q Neural network model as well as the environment for it to work in. I was, furthermore, also in charge of implementing different versions and improvements to the model and algorithms, which had been made over the years in other research papers. I also took part in designing and implementing the third model.
Application Project Description In a team of six other developers, I worked on developing a digital application for planning meetings and lessons for a class of students at a driving school. By being in close contact with the client, we ensured that his problems, needs and opinions were evaluated throughout the entire project. The solution could, given a class of students and the type of driving license, plan and schedule the required lessons needed for each student, give an overview of lessons executed, as well as work as a communication platform between teacher and student.
Function & Responsibility During the project, I worked on planning and designing the UI and components to accommodate the needs of the client. In addition to this, I also built the backend used for functionality, and established the communication with the database.
Designing, defining and implementing a programming language Project Description In colaboration with five other developers, we designed an developed our own programming language, in relation to which we defined its use case, semantics and syntax. A simple programming language capable of defining functionality for digital board games was designed and an interpreter was developed using Java, which interpreted the language into Java code to be compiled.
Function & Responsibility During the project, I took part in planning and designing the language and its use cases as well as in designing and defining the semantics for its syntax. I also implemented various parts of the language functionality and the interpreter used for interpreting it into Java.
Project Application Customer: Project Description Self-employed In cooperation with five other developers, I worked on developing a digital application for recommending tourist attractions in Denmark to users. A MySQL database was Technologies: populated with attractions using Google Places API and users could, based on their MySQL, C# / .NET preferences and location, be recommended a selection of nearby attractions. Using Google Maps API, a route was planned and presented using their selection. Users were able to rate and review their selected attractions.
Function & Responsibility During the project, I took part in setting up and populating the MySQL database using Google Places. Moreover, I was also responsible for developing the backend functionality for the users and for establishing the communication with the database and Google API, used for route planning.
Project Rule mining Customer: Project Description Self-employed Working in a team of four on a scientific project, I implemented and solved some of the complexity problems with temporal association rule mining. Using a large dataset of Technologies: the energy usage of all outlets in a house over multiple years, the data was Python discretized, analyzed and temporal association rules were mined in order to create rules regarding which outlets were likely to be used at which time and time interval.
Additionaly, a dataset of traffic collisions and weather in Manhattan was also examined for the purpose of finding rules correlating between the two datasets.
Function & Responsibility As part of the project, I was in charge of discretizing the data into a more useful format as well as for analyzing the data and the rule mining results. I also took part in designing and implementing various algorithms used for rule mining with the aim of reducing complexity.
Project Explainability of black-box machine learning models Customer: Project Description Self-employed In a team of three other developers, I researched the field of Machine Learning Explainability and the different ways of interpreting and explaining complex machine Technologies: learning models. Working together with a material scientist from DTU, we designed and Python developed an explainability method for their Machine Learning model, which was capable of explaining which features were more important as well as of reducing the feature space to a minimal space, without the loss of performance.
Function & Responsibility Throughout the duration of the project, I was sat as a developer working on all aspects of the project, where I researched, designed and implemented various algorithms for the Explainability method and space reduction of the feature space. In addition to this, I was also responsible for reimplementing the Machine Learning model in TensorFlow 2.0 from 1.0. I was also in charge of developing methods for easing the process of modifying and creating datasets, as well as for developing methods for easily displaying the data in graphs.
it-minds.dk +45 71 99 28 88