Konrad Gałuszka

The system was responsible for presenting the physical status of the automated warehouse. One of the system's key challenges was to comprehend the complicated nature of the location and highlight only essential information in the clearest way. Both the user requirements as well as technical aspects were treated as a priority. User feedback was constantly considered during the continuous development and deployment process. High availability, modularity of the codebase, and the possibility to scale horizontally were important factors from the early stages.

Node.js-based application with convenient UI to visualize the configuration of physical hosts and virtual machines. The main responsibility of the app is to help compare different hosts and fix potential issues. The focus is on the presentation layer to make it as customizable as possible. The data model was designed to be very flexible. The modern UI REST API and GraphQL API were exposed for other systems to integrate and the data to be reused.

The main technologies used are Node.js, React, MongoDB, GraphQL.

A Java-based application providing a role-based access control for system logs. The fully automated approach saved significant amounts of time previously spent supporting individual users with their access requests. The application was containerized and hosted in the OpenShift platform taking full advantage of multiple OpenShift/Kubernetes features.

A Java-based application that was responsible for the collection and presenting of application metrics. It allows users to send any custom metrics defined by the source app and display them on highly customizable dashboards, as well as create comprehensive alerts.

The system consists of multiple components such as a metrics aggregator, alerts engine, dashboard page, and HTTP gateway. Such distribution of roles allows the freedom to scale, depending on needs. It is prepared to run in a containerized environment. The application is highly optimized to be resource-efficient where direct memory is used alongside Heap. Meta-information for metrics is stored separately to achieve a highly normalized model of data. Older series points are down-sampled for better memory usage.

An application providing insight into team performance indicators. The key responsibility of the dashboard is to present the data in a smart way allowing users to compare teams and measure improvements over time. The data is presented on an attractive UI consisting of modern-looking UI components. The presented view allows for the flexible filtering of data. Multiple metrics are available through different back ends that query data from various resources (like change management systems, Jira, and Git), aggregate and cache it to improve performance and minimize the impact on other systems.

Used technologies are React, Node.js, Java, Git, Jira, and Prometheus.

A Java-based system for automation of work for public department workers. The system was responsible for handling official communication between employees of the office and entrepreneurs. One of the key challenges in the application was to authorize users (using officially verified tools), reflect public department structure for the whole country, and provide dynamic templates for numerous forms being applied by users. Reach UI was provided to meet the strict requirements of the officials and implemented with the Java Server Faces framework. The back end side was written in Java and Spring.

Monitoring platform for applications deployed on an Openshift and Kubernetes-based environment. Every application could integrate with monitoring tools and access logs, metrics, and traces infrastructure with very little effort. The work required deploying and configuring tools such as EFK stack (Elasticsearch, Fluentd, Kibana), Prometheus, Jaeger, Thanos, and Grafana. Also, Spring Boot starters were maintained to help integrating Java applications. Work also required enhancements to Golang-based projects such as Grafana to meet company requirements.