hrzn.ch

What It Is

hrzn.ch is the personal homepage of Julien Herzen. It is closer to an academic and engineering portfolio than a commercial product website. The site outlines his experience at Isomorphic Labs, Unit8, Swisscom, EPFL, and other organizations, while showcasing projects in machine learning, time series forecasting, generative models, wireless network resource allocation, political data mining, and more.

Core Features

The most valuable part of the site is its index of projects and papers. Darts is a Python machine learning library for time series, emphasizing a unified API, probabilistic forecasting, multi-series training, and modern deep learning models. VeGANs focuses on experiments with GANs and generative models in PyTorch. Beyond these, the site also collects research projects such as Swisscom Mobility Insights, Mining Democracy, wireless network spectrum allocation, and graph-embedding-based routing, with paper PDFs, BibTeX entries, abstracts, and related links.

Pricing

The site itself is freely accessible, with no registration, subscription, or paywall. Darts and VeGANs, mentioned in the content, can be installed via pip and link to GitHub, indicating open-source projects. However, the website itself does not offer commercial support, hosted services, or paid plans.

Pros and Cons

The main strengths are its tight focus and complete evidence trail across the author’s résumé, projects, and publications, making it useful for quickly assessing his technical background in machine learning and network systems. Paper entries include abstracts and BibTeX, which is convenient for academic citation. The drawbacks are that this is not a full product documentation site: the pages are fairly static, with no search, tag filtering, Chinese-language explanation, or beginner-oriented tutorial system. Some external links, such as GitHub, Medium, and paper PDFs, may be unstable to access from mainland China depending on the network environment.

Who It’s For

It is suitable for time series forecasting developers, machine learning researchers, data science teams, network protocol researchers, and recruiters or potential collaborators who want to understand the author’s background. If your goal is to directly purchase cloud services, SaaS tools, or courses, this website is not a good fit.

Access from China

The main domain is a Swiss .ch personal site and should theoretically be directly accessible. However, its key external links include GitHub, Medium, JMLR, EPFL, and other sites, so the access experience from mainland China may be inconsistent. Overall, it should be considered “partially restricted.”