Scientific Software Development for Research and Education

Quellcode 360 GmbH develops custom software solutions for science, research and data-driven projects. Our specialization in scientific software stems from deep roots in academic research – particularly in astronomy, astrophysics and high-performance data analysis.

This focus is backed by the academic experience of Dr. Johannes Puschnig, a former researcher in astrophysics and radio astronomy. His work involved processing and analyzing massive datasets from telescopes and experiments, using techniques such as spectral analysis, Fourier transforms (FFT), and statistical modeling.

What We Build

• Full-Stack Web Development (Custom Platforms)
  We build web platforms tailored precisely to your needs. Our full-stack approach covers both backend development (logic, databases, APIs) and frontend implementation (design, UI, interactivity). On the backend we use proven Python frameworks like Flask and Django; on the frontend we rely on modern JavaScript technologies such as React. This enables us to deliver powerful, user-friendly web applications – from internal tools to large-scale customer portals.
• Custom Tools for Data Analysis and Model Tuning
  We create customized tools for analyzing large and complex datasets – interactive, efficient, and scientifically robust. Whether time series, simulation outputs, or statistical models: our tools help you extract meaningful insights and present them clearly.
• Big Data Pipelines for Natural Sciences
  For data-intensive research projects, we build scalable pipelines that automatically process, filter, and analyze raw data from experiments or simulations. Technologies such as Python, Dask, Apache Arrow, or Redis form the backbone of these solutions.
• Web-Based Platforms for Research and Scientific Collaboration
  We design browser-based applications to support collaborative research workflows. These include data sharing environments, joint analysis tools, and user management systems for interdisciplinary teams.
• Machine Learning and AI Model Training
  From classification and prediction to pattern recognition: we develop and train AI models based on your data – ranging from traditional ML algorithms to modern deep learning networks. We also advise on model selection, training strategies, and validation methods.
• Interactive Tools for Scientific Visualization
  Scientific data only unfolds its potential when visualized effectively. We build interactive dashboards, spectral analysis tools, 2D/3D plots and map-based views using libraries such as Bokeh, Plotly, Leaflet or Three.js – directly usable in the browser.

Technical Stack

• Languages: Python, C, C++, JavaScript, SQL
• Libraries: numpy, numba, pandas, scipy, matplotlib, astropy
• Frameworks: Flask, Django, React, Node.js
• Scientific tools: MCMC (emcee), FFT, NLTE models, HDF5, FITS, PySME
• Deployment: Docker, GitHub Actions, REST APIs

Read more about our complete tech stack (in German only).

Our tools are used in active research, including large-scale parameter estimation and chemical abundance analysis. One example is webSME, a modern full-stack platform for stellar spectroscopy - developed as part of academic collaborations and now freely available.

With our experience in transforming scientific methods into robust, user-friendly software, we support universities, research teams and science-driven companies in making their data accessible and insightful.

Use Case: webSME

• Online analysis tool for high-resolution stellar spectra
• Built with Python, Flask, Celery and Redis
• Includes MCMC (Monte Carlo Markov Chain) sampling
• Optimized for large wavelength ranges

We also offer consulting and development partnerships to create new tools for research fields such as Earth observation, medical imaging, and time-series analysis.

Wie können wir helfen?

Jetzt 50 % Neukundenrabatt sichern! Den Rabattcode erhalten Sie direkt per E-Mail.