Small video editor software for video editing with filters

A small editor to import, apply different filters and export video files. Made with Qt Designer, PyQt6, OpenCV, PyInstaller and Numpy.

Cloud infrasturucture monitoring system

Implementation of a monitoring system using Grafana, Prometheus, Promtail, and Loki to track the performance and errors of the company's own IoT asset management platform. The development team receives notifications via MS Teams whenever an alert is triggered.

Company: VONOVIA SE

Backend service for data integration

Design and implementation of a backend service that enables the integration of company owned solar energy assets data from a third-party platform into the company's asset management platform.

Company: VONOVIA SE

Backend and templates for cloud platform's notification system

Backend and unittest development of a notification system and an array of notification templates, enabling the seamless delivery of thousands of notifications per day.

This event-driven service not only ensures efficient communication, but also delivers personalized and timely notifications across multiple channels, improving user engagement and productivity.

The platform is used to monitor the company's field assets in real time. Notifications are generated automatically in case of an asset failure or by user action.

Company: VONOVIA SE
More about the platform: VONOVIA press release

Backend service for cloud platform user configurations

Backend and unittest development of a global configuration service that allows users to customise the functionality and preferences of the cloud platform, such as notification types, notification channels, geographical areas of interest of assets and language settings.

Company: VONOVIA SE
More about the platform: VONOVIA press release

An online quotation and booking system

An online quotation and booking system for a beeswax processing plant in southern Chile (Patagonia). Developed with Django, Celery, HTMX and Jquery.

Company: APISUR

Planning, design and construction of an automated test bench for continuous integration

A test bench for continuous integration. Before a new release for a specific part of the collaborative robot (cobot) is completed, the software undergoes automated testing using the test bench. The objective is to enhance the final product's quality by identifying and addressing all potential errors before deploying the code. Multiple tests are conducted, carefully scrutinizing different components.

Company: Yuanda Robotics GmbH (Now voraus robotik)

IoT device for monitoring and optimizing forestry workflows

This IoT device tracks parameters such as elapsed time, position, distance, speed, and the number of load transport work cycles. It utilizes an embedded Linux device equipped with an integrated SIM7600 modem for acquiring GPS data. The device connects remotely to a Django Application using MQTT over TLS for data transmission. Within the application, users can view the data in a table format and track the position in real-time on a map. Furthermore, the data can be effortlessly exported in multiple formats.

Basic diagram: click to open.

Plug-and-Play Temperature Controller

A fully digital, robust, easy to use and quick to install plug-and-play temperature controller for the honey processing industry. It has been specially designed for flexible use in various applications where temperature regulation is required.

Documentation: click to see.

Some Circuit Layout Designs

Printed circuit boards designed with KiCAD. SMD components are soldered by the manufacturer. Missing components are assembled manually.

Some Membrane keypads

Membrane keyboards. I design them with GIMP and Inkscape.

Some Machine Learning and IoT Work

A Softsensor, a Linux-based embedded system with a connected vibration sensor, is strategically placed on a machine or process for continuous monitoring. This embedded system runs a specialized program to collect data for learning purposes. The collected data is then seamlessly transferred to the Amazon cloud (AWS), where an inference model is generated.

This inference model is a crucial tool for real-time monitoring and is complemented by various cloud services to enhance security and availability. In the event of a detected failure, the system automatically triggers instant notifications to technicians, ensuring swift response and action.

Github: click to see.
Image: SSV Software Systems GmbH

An Open Source Library

PyDSlog is an open-source library designed to streamline the collection of data essential for training supervised machine learning models. It simplifies the process by providing a mechanism to gather data in the form of a stream. This data stream can be directed straight to an inference model for real-time analysis or used to generate CSV files for further processing and model training.

Data collection is a vital step in the development of machine learning models, enabling them to learn and make accurate predictions. PyDSlog makes this process more efficient, making it a valuable tool for machine learning practitioners.

Github: click to see.
PyPI: click to see.
Image: SSV Software Systems GmbH