Hi Raphael, could you tell us a bit about yourself in a few words?

I’m Raphael, and I’m a data engineer at Flowlity. I’m responsible for the data pipeline of our SaaS solution. I am responsible for the quality of data on the different layers of our product and the robustness of the different transformation processes, such as connections to the customer’s ERP, raw data integration, synchronization and standardization, analysis/BI, machine learning and optimization models, application and cached data, and exporting and integrating results into the customer’s ERP. I enjoy designing, implementing, and maintaining data architectures for innovative solutions. I have a dual honors degree in Mathematics and Statistics from the Sorbonne and Computer Science from 42 France.

I’ve been in charge of the Data team at Flowlity for a year. We offer a data product and our ambition is to be integrated into all the links in the supply chain. There are many solutions, systems and tools in the supply chain ecosystem, and it is a real challenge to enrich these different solutions. It is a team that will grow because of its strategic importance.

I started my career as a data scientist and researcher, but I really enjoy the tech and production side of the data engineer role. At Flowlity, I continue to support the research team, but I also have the chance to work with devops, frontend developers, sales and customer success!

In addition to these activities, I have the pleasure of teaching on the Master 2 – Statistics program at the Sorbonne, on a machine learning topic that I particularly enjoy: Reinforcement learning.

If you had to describe your job in 3 words:

  • Development: I work on new features and on the improvement and orchestration of the data pipeline: backend, model ML/optim, devops, etc. We organize ourselves in a 2-week sprint, with priority tickets at different levels of our stack (python, sql, airflow, k8s, azure, etc.), with regular releases on our different environments.
  • Analysis: I have a lot of interaction with the customer. Before integration, we need to understand their IT and data modeling and agree on the integration steps (security, connection, protocols and mapping). In collaboration with the delivery team, I analyze (notebook, BI) the raw data to validate the customer workflow and integration specs (mapping, filter, …).
  • Resilience: Even with the best ML models in the world, if the data is no good, the results will be no good. The data is never perfect, and this can be for many different reasons, whether they are technical, human, internal or external. You have to stay vigilant and be constantly alert to continue to improve quality and prevent potential errors.

What is your typical day like?

I start my day at 9am, and I look at the logs to see if there are any alerts or bugs. When there are errors, I’ll start by fixing them. This is always my priority.

Then, I focus on the clients (data, queries, etc.), and then I’ll look at the features that need to be developed or the code to refactor. Then I eat (laughs). And then I carry on with this work into the afternoon.

I also have a lot of meetings with customers and/or synchronization points with other teams to understand each others’ needs. It is important to understand when a change is happening so that we can roll it out properly.

What sort of interaction do you have with customers?

I interact with them on several levels:

  • Potential clients: even before we sign a service agreement, I will meet with the client to assess feasibility from a data perspective. I will talk with their IT team and describe our infrastructure, security, data, etc.
  • Client undergoing integration: At this stage, we write and validate the specs for data integration and infrastructure connection protocols with their IT team, etc.
  • Live client: integrate new data and model changes, fix bugs and develop new features, etc.

I really work closely with the delivery/customer success teams on these matters.

Why Flowlity?

Before Flowlity, I was at Eau de Paris. It was a great company with very interesting issues and they were looking to go digital. I really enjoyed that experience, but I wanted to return to a start-up environment, something more agile and international.

I knew Arthur (Head of research) and he convinced me to join.

What convinced me was the ability to deliver. The solution that has been developed in just two years is really impressive. And I really like the organization that goes with it, in terms of an agile methodology. I’m here for the team. There are some people here with solid experience and everyone is committed. There is real solidarity and good communication in the team.

We live with our customers too. We constantly follow their needs and adapt, and I really like that.

I didn’t know much about supply chain management, but since joining Flowlity, I can see now that it is everywhere! There are several types of customers, but they all have important related problems. And I believe that Flowlity has the ability and expertise to provide a real solution that can make life much easier for these companies!