Samir Sekkat

Freelance senior engineer and technical lead, Berlin

For twenty-five years I have built production systems and led the teams behind them, across startups, scale-ups, and enterprises. I take on the work that needs real depth: designing the core of a new system, making unstable systems reliable, or stepping in to steady a team through a difficult stretch. I work hands-on, close to the team, and I care that the outcome is right.

How I can help

Make it reliable

My particular strength is taking systems that have become unstable, slow, or opaque and making them stable and dependable: eliminating cascading failures, building the observability that was never there, and delivering performance gains measured in multiples, not percentages. I hold a high bar on testing and clean architecture, because that is what reliability actually requires.

Build the hard part, together

The core of a new system is often where the real thinking happens. I work closely with founders and teams to turn an evolving vision into something concrete that the rest of the product can rest on, and to advise on direction as it takes shape. It is collaborative, hands-on work, and getting the foundation right is what lets everything built on top of it move quickly and safely.

Build and lift the team

I have built distributed teams across countries, stabilised an organisation in crisis and grown it to market leadership, and helped early-stage startups hire and grow the people who carry the work forward. Decades of coaching practice mean I raise the teams I work with, through pair programming, review, and candour even when the honest thing is hard to say.

On AI

I use AI intensively, but as the opposite of hands-off code generation. Mostly I work with it in pair-programming mode: a thinking partner that proposes, argues, and widens the search, while I stay in control and understand the source code that ships. I also lean on it to run quick, disposable experiments that probe a problem cheaply. Explore fast, then build with care. On any horizon longer than a year, doing it right the first time is the faster path; it is what spares you the rewrites that quietly cost most teams their speed.

Graph data, and why it matters now

Long before it was fashionable, I built systems on graph databases: designing RDF and SPARQL data architectures, migrating between engines for real performance gains, and bringing teams up to speed on a way of modelling data that few engineers have touched. That experience matters more now, not less. As AI systems grow, the hard problem is no longer generating text; it is giving models structured, trustworthy, connected context to reason over. Graphs are how you represent knowledge so that both people and machines can navigate it. I help teams model their domain as a graph, and connect that foundation to the AI systems that increasingly depend on it.

About

I am drawn to a single question that runs through everything I do: how does complexity arise, in software systems and in human behaviour, and how do you simplify the one while working patiently with the other? It is why I go deep technically and why I care as much about the people as the code. I work remotely from Berlin, in French, German, and English, and I stay open to the experience and wisdom of others, knowing how small a slice of it any one person holds.

Let’s work together

If you have a system to build, one that needs untangling, a team to grow, or a problem you are still trying to frame, write to me at sse@sekkat.eu and tell me what you are working on.