Benoit De Menthière

Data scientist

Moves to Paris

  • 48.8546
  • 2.34771
Propose a project The project will only begin when you accept Benoit's quote.
Propose a project The project will only begin when you accept Benoit's quote.

Location and geographical scope

Location
Paris, France
Can work in your office at
  • Paris and 50km around

Preferences

Business sector
  • Aviation & Aerospace
  • Automobile
  • Energy & Utilities
  • Environment
  • Mechanical or Industrial Engineering
+2 autres

Verifications

Influence

Github

Github : bensdm bensdm
  • 0 Followers
  • 8 Repos
  • 0 Gists

Languages

Categories

Skills (21)

Benoit in a few words

Diplômé d'un mastère spécialisé Big data et Machine learning à Telecom Paristech après avoir suivi une formation d'ingénieur généraliste aux Arts et Metiers qui m'a permis d'acquérir de bonnes connaissances métiers.

Je peux vous aider à:
- mettre en valeur vos données (nettoyage, pipelines...)
- segmenter vos données
- prédire des valeurs futures (industrie...)
- visualiser et identifier des patterns
- détecter des anomalies (log, comportements, outliers...)
- extraire des données du web (scraping)
- extraire des données de documents (OCR)
- faire du traitement d'images (computer vision)

Experience

Schlumberger - Schlumberger

Energy & Utilities

Data Scientist

Houston, États-Unis d'Amérique

July 2019 - March 2020

Cement thickening time prediction:
• Optimization of the preprocessing pipeline: 8 times faster
• Build dashboard to control data quality
• Clustering of experiments and outlier detection
• Build regression model with confidence estimation: reduce errors
by 25%
• Deploy user-friendly webapp dividing the time needed to find the
optimal cement recipe by 30
Maintenance cost prediction:
• Build a constrained linear regression model
Subsurface fiber optics responses prediction:
• Unsupervised pretraining using infoVAE on 30M data points
• Expected savings $1M/well

Bearing point

Energy & Utilities

Data Scientist

Paris, France

September 2018 - June 2019

Optimization of the energy consumption of a silicon furnace:
• Data analysis and visualisation to identify causes of decline in silicon
production
• ARMA model to anticipate these drops

SopraSteria

Defense & Military

Data scientist

Courbevoie, France

February 2018 - August 2018

Trajectory clustering via deep learning representation using auto-
encoder:
• Fine tuning parameters
• Visualization
Optical character recognition and entities extractions on invoices and
medical prescriptions:
• Computer Vision with OpenCV
• Natural language processing with NLTK
• Deployment with docker

1 external recommendation

Education

Certifications