Role
Full-stack web engineering and analytics at BYND Network
I joined BYND Network as a web developer in May 2023. Since then I have worked on the team shipping full-stack websites for more than 15 of the agency's clients, from the interface people click on to the back end and data behind it. Shopify storefronts and custom Shopify apps are part of what I build there, not all of it. I also built performance dashboards for internal teams and clients, and used data analysis to improve conversion by 30%.
The work
BYND Network is a web agency, and it is where my engineering first had real customers attached to it. Since joining in 2023 I have been part of the team that ships full-stack websites for more than 15 clients. On those projects I work across the front end, the back end, the data and the deployment, and each site is delivered as a working product, not a mockup.
Shopify is one of the tools I use there, not my whole job. When a client sells online, I build and customise their storefront with Liquid, JavaScript and the Shopify API. When their operations do not fit what the platform offers out of the box, I write custom Shopify apps. When a client needs something Shopify was never meant to do, the site gets built from scratch, and I build my share of it end to end.
Dashboards people actually open
Client sites generate more data than anyone reads. I built performance dashboards for internal teams and for clients so KPI tracking stopped being a monthly export and became something a team could look at and act on. The requirement was never sophistication. It was that the number on the screen is the number someone can make a decision with.
Analysis with a result attached
Applying data analysis to how visitors actually behaved on a client's site improved conversion by 30%. That figure is the clearest argument I can make for why I work on both sides of the problem: the modelling is only worth as much as the business decision it changes.
Why it belongs in an ML portfolio
Three years of shipping to clients teaches the parts of machine learning engineering that research does not: that requirements move, that a system has to be maintainable by someone else, and that insight has to be delivered in a form a non-technical stakeholder can act on.
What BYND taught me
I'm Ralph, a machine learning engineer and data scientist based in Paris. At BYND I learned to own my part of a project from end to end. On a client site that meant everything from the database to the button someone actually presses. I bring the same habit to models, following them from the raw data to the decision they are supposed to change. Working at an agency taught me to do that against real deadlines, for clients who notice when something breaks.
At BYND the next project is always waiting, so whatever I hand over has to hold up without me. I write code a teammate can pick up without calling me, and I built dashboards clients open without a reminder and numbers I can explain in plain language. If something I built only makes sense to me, I don't count it as finished.
On the team I am usually the one asking why. Why does this page convert worse on mobile? Why did that metric drop on a Tuesday? The 30% conversion lift started with a question like that and some patience with the data. Building websites at BYND taught me to ship. Data science taught me to check whether what I shipped actually worked. I want to keep doing both.