Engineering the flow of data
High-performance data platforms, streaming pipelines, and production-grade AI infrastructure.
Worked with
Capabilities
Data Engineering
Large-scale data platforms, lakehouse pipelines, and distributed processing systems.
Streaming Systems
Real-time event processing, stateful pipelines and resilient dataflows.
AI Infrastructure
Production ML pipelines, evaluation systems and scalable AI platforms.
Cloud Architecture
High-performance distributed architectures optimized for reliability and scale.
About
Laminar Dataflow builds high-performance data systems at the intersection of data engineering, cloud computing and artificial intelligence. We specialize in distributed architectures, streaming pipelines and scalable AI infrastructure.
Founder
Thomas Rouch
Data Engineer & Architect — Spark, Scala, Databricks, Azure
20+ years in software engineering, 14 of them in data. I started in embedded and real-time systems in C/C++, which left me with a lasting taste for performance, memory and the cost of every millisecond.
Today I design and build lakehouse platforms and large-scale Spark pipelines for airline, banking, telecom and industrial companies: incremental ingestion, Data Vault 2.0 and medallion modelling, performance tuning, and migrations of legacy stacks to Databricks or Snowflake.
I usually join a team as a hands-on tech lead or expert, or run short audits of an existing platform. On my own time I explore neuroevolution: a GPU-accelerated genetic algorithm that trains a small neural network to land a Mars module on CodinGame.