Streaming & AI Solutions Architect
Fifteen years in production. Now an independent practice.
I’m Aurélien. For 15 years I built mission-critical streaming and AI systems inside banks, gaming platforms and public-sector data — as a senior engineer, hands-on. I’ve opened Scalefine to do the same work directly for the teams that need it: diagnose, rescue, and industrialise the systems that can’t afford to fail.
100K+ events/min · 97% accuracy · 18× speedups · €10M+/day in production
The thread
Make the hard machinery disappear — so lean teams hit harder.
It started with a generalist engineering degree at the École des Mines — an innovation option, a Technologies specialty, and an innovation prize from Legrand along the way. The career began on Paris trading floors and fintech, then went international — Prague, Germany, now remote from a Singapore base. I founded my own marketplace startup somewhere in there, and the builder’s itch never left: today it’s an independent practice — with a robotics venture taking shape on the side.
The constant through all of it: take the heavy, real-time, AI-heavy plumbing off people’s plates. I build data streaming and AI automations that turn slow, manual, error-prone work into systems that run themselves — so a small team can punch well above its weight, and spend its energy on what actually matters.
Track record
Three domains, one common thread: production.
Real-time data and AI, shipped and kept running under regulated, high-velocity conditions — not demos.
Banking · 3 years
Trading floor
Derivatives automation at €10M+/day on the live floor — rules-engine & AI solver inside a Java/Spring platform, audit trail intact, at Société Générale CIB (Paris).
Gaming · 5 years
Streaming analytics
Real-time Kafka/Scala microservices at 100K+ events/min, sub-300ms, plus production graph-AI / RAG solvers at 97% accuracy, at ESL (Germany).
Public sector · IoT & security
Critical data systems
High-throughput, security-sensitive data work for Thales, the European Parliament and the Schengen SIS-II — including an Oracle→Cassandra migration at 18× throughput.
Reference
In their words.
“His capacity to quickly understand complex systems and develop creative, unconventional solutions was remarkable. His resilience during critical incidents was particularly noteworthy.”
ESL FACEIT Group · official Letter of Reference (Oct 2025)
Full reference letter and public LinkedIn recommendations available on request.
How I work
Remote-first, clean, accountable.
Engagement
Direct B2B via Scalefine Pte Ltd (Singapore) · clean EUR/USD invoicing · Outside-IR35 · no payment intermediary.
Availability
Remote-first (CET overlap) · Day-1 ready in ~2 weeks · references provable in 2 clicks.
Method
Vendor-neutral, cost-conscious, Pareto engineering — fix, optimise and industrialise, then hand over cleanly.
Range
From the event stream to model behaviour — streaming, expert systems and production AI, end-to-end.
Stack
What I build with.
Apache Kafka · StreamsFlink · SparkScala · Java · PythonAkka · Cats EffectgRPC · GraphQLGenerative AI · LLMRAG · vector DB · FAISSLangGraph · multi-agentrule-engine / AI solverSpring Boot · FastAPIAWS · EKSKubernetes · Docker · TerraformCassandra · PostgreSQLPrometheus · Grafana
Credentials
Education & certifications.
- 🎓 Mines d’Alès — Engineering (Software + Innovation)
- ☁️ AWS Certified Solutions Architect – AssociateCredly-verifiable
- 📊 Udacity — Data Engineering Nanodegree
- 🤖 Udacity — AI & Deep Learning Nanodegree
Off the clock
What’s next.
Alongside the practice, I’m building a small robotics venture on the side — the same instinct that drives the rest: take something hard from prototype to a thing that actually works. Remote-first across Europe and Asia.
Let’s talk
A 20-minute call to see if it fits.
No prep, no slides. We figure out if there’s a fit — and if not, you walk away with one concrete suggestion. That’s the offer.