Stéphane Couvreur

Lisbon, Portugal · [email protected]

I am a Data Engineer at Cloudflare.

I have a passion for big data, real-time processing and distributed systems.

Experience

Senior Data Engineer

Apr 2022 – Present
Cloudflare, Lisbon
  • Migrated Apache Spark ingestion pipelines to Go improving peak throughputs from ~85k to 1.2M rows/s from Postgres, reading >4B rows/day.
  • Reduced daily Postgres ingestion memory consumption 94% and ingestion time from 15h to 1h20m using b-tree index histograms, relieving constrained CPU and memory in core datacenter.

Data Engineer

Jan 2021 – Mar 2022
Cloudflare, Lisbon
  • Developed data pipelines in Apache Spark with throughputs of >30B rows and >5TB/day.
  • Aggregated ClickHouse traffic and Prometheus metrics data from core and edge data centers.
  • Built data lineage sensing tools in Airflow, reducing data lineage incidents by 37%. Reduced MTTR and number of incidents by >50% QoQ, saving a 22-person team >800 man-hours per quarter.

Data Scientist

Oct 2018 – Dec 2020
Feedzai, Lisbon
  • Developed ML models detecting transaction fraud with 78% $ recall at 1% FPR in Apache Spark.
  • Built data pipelines on Apache Hadoop clusters to train models on >1B rows of data.
  • Supported production systems with throughputs of >2k events/s and <200ms latency at 99.999%.

Research Assistant

Jun 2017 – Sep 2018
Great Ormond Street Hospital for Children, London
  • Collaborated with engineers and clinicians building cardiac MRI segmentation pipelines.
  • Implemented a whole-heart segmentation workflow achieving an AUC of 0.752 using 3D CNNs.
  • Secured 2 successful grants from NVIDIA and Medtronic.

Education

UCL, London

Sep 2013 – Jun 2017
Master’s Degree, Mechanical Engineering

Thesis in Computer Science and Medical Imaging.

Skills

Programming Languages & Tools
  • Apache Spark
  • Apache Hadoop
  • Apache Airflow
Certifications
  • Certified Scrum Master 2018