Mission

Our mission is to foster a culture of data-driven decision-making within organizations. We achieve this by providing comprehensive training in data analysis and creating innovative software tools that promote reproducible research.

Send a message over LinkedIn

Services

Paul Rougieux partners with clients and their teams on the following 3 types of services:

  • Development. We develop customized software solutions that convert your raw data into a format suitable for your modelling applications. These solutions are designed to be integrated into your existing processes and to automate your workflows.

  • Pair programming sessions. Half an hour to 2 hours during which I provide technical support to a technician or engineer working on their data processing needs. It could be building a data transformation pipeline or implementing a model. I review the code to improve reproducibility.

  • Analytics team building. When expertise is separated between domain experts on the one side and programmers on the other side, there is a friction between data science models and the users of those tools. I aim to ease that friction by training the experts so that they are able to program their software tools themselves. Throughout the years, domain experts have learned mental models that are an internal representation of their complex systems. For their analysis needs they have learned to use spreadsheets and special purpose analytics software. Domain experts are best equipped to build functions and objects that fulfil their analysis needs. Together we aim to structure a software package in a way that it becomes natural to interact with its functions and objects for a user with specific domain knowledge. For this purpose, I train team members to use the tools of reproducible research such as version control, programming and software packaging. In parallel, I train engineers to build a data pipeline that can be scaled up and reproducible.

Why build-up analytic teams?

In the analytics game, different roles have different questions :

  • For managers : do you want to use data analytics to transform your organisation? If yes, have you thought about empowering all your colleagues to create data science tools themselves?

  • For programmers : when talking to a domain expert, have you had the impression of talking to a tree?

  • For domain experts, technicians, engineers, doctors: are you frustrated with the current way the analytics tools are setup in your company?

When technicians are hungry enough to learn how to create and automate analytic tools, we have a very good starting point. The software they will draft is not only useful to their analysis immediately, it also becomes a communication tool to describe their mental model to other analysts and programmers in the organisation. Indeed complex systems are difficult to encode properly in software by a programmer who doesn’t have the right mental model. When domain expert learn to write software themselves, it becomes not only an analysis tool that they can use immediately for their own needs, but also a very good vehicle to communicate about complex systems and their representation inside models. Now that programmers have source code showing what analysis are required by the experts. They can take that source code to enhance it for reproducibility and to scale it up as an operational model. When management is leading their team so strongly that they are ready to train their technical staff in the craft of creating analytics software. That is where we come to propose our pair programming and analytics team building services.