My work ethic is based on two main aspects: sobriety and sharing. I conceive of myself as a craftsperson in the sense that I personally handle all stages of my production (from dialogue with the client to commercialization). I also claim a craft activity because of its aesthetic character and the care I bring to my work. I rely on the notion of convivial tool developed by Ivan Illich in the 1970s.
Sobriety
Material and Energy Sobriety
I attach great importance to the choice of equipment I use in my professional and personal life. Thus, I favor sober, ethical and sustainable1 computer equipment. I have signed the ecological manifesto of IT professionals. I am proud to work on the same computer since the creation of my company. I regularly audit my websites to reduce their ecological impact.
Analysis Sobriety
I believe in the beauty of simplicity and I try to keep this objective of sobriety in mind when conducting analyses. For example, within the trade-off between (i) performance of statistical methods and (ii) level of interpretability of results, I favor the latter2.
Sharing and Transparency
I offer my services at a controlled cost by prioritizing transparency and integrity.
Sharing Knowledge and Tools
I commit to providing you with all annotated analysis scripts and pipelines3 and to answering your questions on this subject without time limit. My objective is to provide you with maximum autonomy in the use of the tools I use/develop. Part of my work therefore consists of training you in these tools.
The Free Software Philosophy
I work under Linux4 and use almost exclusively free software. I participate in this ecosystem on my own scale by developing and maintaining R packages.
Free is not free of charge: I donate 1% of my revenue to help develop the free software I use most. This commitment joins that of other actors involved in the Copie publique project.
1. For example, I use an Eizo screen,
computers running Linux assembled in France and easily
modifiable, and a phone running
/e/os (an Android
distribution that allows you to do without Google).
[↩]
2. I do not deny the need for complex forms of analysis in
certain situations (e.g. machine learning techniques),
but I openly opt for an attitude oriented towards
interpretability of analyses and understanding of results
for my clients.
[↩]
3. I favor simple and universal file formats such as
(R)markdown, svg, odt and git for example.
[↩]
4. Mainly with the Kubuntu distribution.
[↩]