17.04.2024 aktualisiert
60 % verfügbar
Principal Solution Architect, Full-Stack-AI-Developer & Dev-Team-Upgrader
Mauer, Deutschland
Deutschland
B.Sc. WirtschaftsinformatikSkills
QunitTest-SoftwareClean CodeGulp.JsTrainingsdatenHTMLJavaJavascriptPHP3D Modellierung3D PrintingAbrechnungssystemeAPIsAgile MethodologieKünstliche IntelligenzData AnalysisAndroidAndroid Application DevelopmentApache HTTP ServerApache TomcatAnwendungsserverGeschäftslogikArchitekturmusterComputer VisionConfluenceJiraHtml5AuthenticationsTest AutomationUnit TestingMicrosoft AzureLAMPBig DataC#CSSCloud ComputingCode GenerationSoftware QualityCUDADatenbankenContinuous IntegrationData IntegrationDatenverbundDatenvisualisierungDebian LinuxLinuxExtensible Application Markup LanguageFfmpegGithubGlassfish 3Google GuiceGradleHadoop Distributed File SystemHardware-DesignWeb-PortaleSwing (Java)JsonWildflyApache JmeterJavaserver FacesJunitPythonMachine LearningMemcachedMongodbMysqlNode.JsOauthE-AccountingOpencvOpen SourcePair ProgrammingPerformance-TuningScrumSoftware ArchitectureRaspberry PiTensorflowAngular 2+SwaggerResponsive WebdesignRSA (Cryptosystem)ScreenshotsMicrosoft SharepointSingle Sign-OnSoftwareentwicklungSoftware SystemsTypescriptUI TestingUMLVaadinWeb ApplikationenWeb Application FrameworkWebseitenApache WicketWindows Media PlayerExtensible Markup LanguageOpenapiDatenverarbeitungPrimefacesModellgetriebene EntwicklungTest-Driven DevelopmentAngular MaterialReactJSRollupExpress.jsIT-ArchitekturGitlabGitFastapiPandasEvent Driven ArchitectureModularizationMatplotlibWebpackAngularJSKubernetesApache KafkaBuild-ProzessWeb TechnologiesFront EndText AnalysisWinkelmesserRequirejsSplunkNetwork ServerMochaJasminAtlassian BambooDockerJenkins
3D printing, 3D design, Accounting Systems, Agile Development, Agile Methods, Android apps, Android SDK, Angular 2, Angular2, Angular Material, Angular, Apache, Apache jMeter, Kafka, Apache Tomcat, Apache Wicket, API, architectural patterns, AI, Bamboo, test automation, automated tests, automated testing, large data, Big Data, Big Data technologies, build process, business logic, C#, CSS, SCSS, Clean Code, Cloud, Code Generator, image recognition, Confluence, CI/CD, Data Analysis, Data Integration, data processing and analysis, Shared data, Data Visualization, data visualizations, Database, Debian, Docker, Online Accounting, Event-Driven Architecture, Express.js, XAML, XML, FFMPEG, FastAPI, Frontend, GIT, GitHub, GitLab, GlassFish, Google Guice, Gradle, Gulp.js, HTML, HTML5, HTML 5, HDFS, hardware integration, user interaction, IT architectures, Jira Agile, Jira, JSON, Jasmine, Java, JavaScript, JSF, Jenkins, jUnit, Kubernetes, LAMP stack, Linux, Machine Learning, Matplotlib, memcached, Azure DevOps, Azure Cloud, Azure, MS Azure, SharePoint, Windows 8, Mocha, model-driven development, Modularization, MongoDB, MySQL, servers, Node.js, NodeJS, CUDA, OAuth2, Open-Source, OpenCV, OpenAPI, PHP, Pair Programming, Pandas, Performance Optimization, PrimeFaces, Protractor, Python, qUnit, RSA, Raspberry Pi, React, RequireJS, responsive web, Responsive Design, Rollup, Screenshot, Scrum, application servers, single sign-on, software architecture, software development, code quality, Software Solutions, Splunk, Swagger, Java Swing, TensorFlow, test-driven development, testing software, text analysis, training data, TypeScript, UML, unit tests, authentication, UI Testing, Vaadin, Web Frameworks, web application, Web Components, web applications, web pages, web UI, web technologies, web technology, WebPack, JBoss, media player
Sprachen
DeutschMutterspracheEnglischgut
Projekthistorie
In January 2019, I began my role as Lead Frontend Developer by addressing performance issues that emerged following the migration of a substantial web application to Angular. This process included optimizing existing functionalities and enhancing load speeds, as well as transitioning from a proprietary, server-heavy framework to Angular 7, which significantly boosted performance.
Following the successful migration, I developed a component library that substantially improved the software's development and maintenance by enabling developers to efficiently use reusable UI components. Additionally, I implemented comprehensive unit and screenshot tests to ensure the application’s quality, enhancing the reliability of updates.
Since 2020, I have taken on the responsibility of advising and team coaching, focusing on the strategic advancement of the application. My activities include providing architectural advice, coaching the development team, and implementing best practices in modern software development.
A particular success was the restructuring of the project into independent sections, which enhanced the application's modularity and technological independence. The adoption of WebComponents and Stencil amplified our technical agility and facilitated the integration of new features.
This role has allowed me to deepen my expertise in areas such as Angular, JavaScript, DevOps, and agile methodologies, while simultaneously developing my strategic leadership skills in a technical environment.
Following the successful migration, I developed a component library that substantially improved the software's development and maintenance by enabling developers to efficiently use reusable UI components. Additionally, I implemented comprehensive unit and screenshot tests to ensure the application’s quality, enhancing the reliability of updates.
Since 2020, I have taken on the responsibility of advising and team coaching, focusing on the strategic advancement of the application. My activities include providing architectural advice, coaching the development team, and implementing best practices in modern software development.
A particular success was the restructuring of the project into independent sections, which enhanced the application's modularity and technological independence. The adoption of WebComponents and Stencil amplified our technical agility and facilitated the integration of new features.
This role has allowed me to deepen my expertise in areas such as Angular, JavaScript, DevOps, and agile methodologies, while simultaneously developing my strategic leadership skills in a technical environment.
This project aimed to develop an application for verifying IDs captured via an app, featuring a two-stage verification process: an automatic AI evaluation followed by manual review. I developed both the AI assessment and the interface for manual review.
The AI assessment utilized image recognition technologies and text analysis to evaluate the IDs. Due to the limited amount of available training data, data augmentation techniques were employed to enhance the model's robustness. Key challenges included identifying optional markings on the ID and verifying name matches.
For IDs with an AI rating below a predetermined threshold, I developed a user-friendly interface that allowed non-developers to manually check and validate the data. This feature was crucial in building trust in the assessment system.
This project not only allowed me to deepen my skills in AI development but also significantly enhanced the efficiency of the verification process by limiting manual reviews to the most problematic cases. The intensive work on data augmentation techniques provided valuable experience in handling insufficient datasets and underscored the importance of effective data utilization in AI-driven processes.
The AI assessment utilized image recognition technologies and text analysis to evaluate the IDs. Due to the limited amount of available training data, data augmentation techniques were employed to enhance the model's robustness. Key challenges included identifying optional markings on the ID and verifying name matches.
For IDs with an AI rating below a predetermined threshold, I developed a user-friendly interface that allowed non-developers to manually check and validate the data. This feature was crucial in building trust in the assessment system.
This project not only allowed me to deepen my skills in AI development but also significantly enhanced the efficiency of the verification process by limiting manual reviews to the most problematic cases. The intensive work on data augmentation techniques provided valuable experience in handling insufficient datasets and underscored the importance of effective data utilization in AI-driven processes.
In this project, I played a key role in developing a web application based on AI-powered predictions. My focus was on the frontend and the sophisticated visualization of results from AI models, utilizing OpenStreetMap for geographical representations and GraphQL for efficient data queries. The quality of my visualizations and the flexible adaptation of presentation parameters made a substantial contribution to the fine-tuning of AI models. This interactive visualization work was crucial in enhancing the accuracy of predictions and demonstrated the critical role that high-quality visualizations play in the success of AI-driven solutions. This project deepened my understanding of the importance of precise data visualizations in the context of modern technologies and reinforced my commitment to advancing in this area.
Portfolio
AI supported UX
In collaboration with a colleague, I developed a showcase demonstrating how to enhance user experience on conventional web pages for data entry by utilizing AI. By creating models from historical inputs that generate suggestions for input fields and flag unlikely entries as potential errors, we significantly improved data quality. We analyzed anonymized data on vehicle valuations to develop a model that predicts the most likely tires for a used car based on previously entered values such as dimensions and weight. The efficiency of our solution was evident as we increased the probability of identifying the correct tire among the top five suggestions from 12% (without AI) to over 50% (with AI). This success illustrates that even with minimal effort, the user-friendliness of large data-entry forms can be substantially enhanced, which is particularly advantageous in areas with extensive data.
AI ID Verification
As a Full-Stack-AI-Developer I realized an application for verifying IDs captured via a software solution, featuring a two-stage verification process: an automatic AI evaluation followed by manual review. I developed both the AI assessment and the interface for manual review. The AI assessment utilized image recognition technologies and text analysis to evaluate the IDs. Due to the limited amount of available training data, data augmentation techniques were employed to enhance the model's robustness. Key challenges included identifying optional markings on the ID and verifying name matches. For IDs with an AI rating below a predetermined threshold, I developed a user-friendly interface that allowed non-developers to manually check and validate the data. This feature was crucial in building trust in the assessment system. This project not only allowed me to deepen my skills in AI development but also significantly enhanced the efficiency of the verification process by limiting manual reviews to the most problematic cases. The intensive work on data augmentation techniques provided valuable experience in handling insufficient datasets and underscored the importance of effective data utilization in AI-driven processes.
Visual Craft-NN
Project Lead | Ongoing Visual Craft-NN is an educational project I initiated to develop a web application that visualizes neural networks, their topology, and intermediate results. This tool is designed to help a wide audience understand complex concepts, training stages, and differences in neural network operations. The application uses TensorFlow.js and Rete.js to provide real-time, interactive visualizations that elucidate how neural networks process data and make decisions. This project aims not only to educate but also to foster greater transparency and accessibility in the field of artificial intelligence. By leveraging cutting-edge web technologies, Visual Craft-NN allows users to manipulate and observe neural network behaviors directly through a user-friendly interface. This initiative is part of my ongoing commitment to demystifying AI technologies and making them accessible to a broader population.