Naoufal Acharki

Curriculum vitae

Naoufal Acharki

Senior data scientist at Air France-KLM and freelance consultant, based in Paris, France, with a PhD in Applied Mathematics from École Polytechnique. I work on experimentation, causal ML and recommender systems for marketing teams.

Download PDF Email LinkedIn

Experience

  1. Nov 2025 – Present Roissy-en-France, France

    Senior Data Scientist, Marketing Operations Research

    Air France-KLM · Commercial Data Officer, airline group

    • Offer Display Recommender, which personalises the products each customer sees.
    • Funnel, traffic and segment analyses to decide which metrics and guardrails each A/B test on the booking flow should use.
    • Data preparation and exploration for the models, then the offline and online evaluation of every experiment, turned into decisions for the business.
    • +0.6% conversion rate, relative lift one-month A/B tests against control, flight selection and offer display
    • +0.5% revenue per visitor, relative lift same experiments; a further upsell test is significant and awaiting deployment
    • €1M+ incremental revenue per month reported impact of the shipped experiment winners
    PythonSQLBigQueryA/B testingRecommender systems
  2. Aug 2025 – Jul 2026 San Francisco, USA (remote)

    Mathematics and Statistics Expert

    Mercor · Talent platform for AI labs

    • Mathematics and statistics problems that state-of-the-art models get wrong and experts can solve, each with a verified solution.
    • Agentic frameworks and expected model behaviour under several constraints, and the evaluation of models against them.
    • A STEM review project, exploring datasets and producing visualisations, diagrams and infographics with the science checked behind each one.
    ProbabilityStatistical inferenceOptimisationLLM evaluation
  3. Jan 2025 – Jul 2025 Remote

    CTO

    Kemba AI · Startup, agentic AI operating system for banks and fintechs

    • Designed, built and launched the MVP of a multi-agent platform for financial services, from the architecture to the deployment.
    • Ran the technical side of the venture. We didn't raise the funding we needed, and the project stopped in July 2025.
    Multi-agent systemsLLMsAWS
  4. Oct 2023 – Oct 2025 Mexico City, Mexico (remote)

    Senior Data Scientist and Machine Learning Engineer, Pipelines and MLOps Lead

    Senzai · Marketing-analytics startup

    • Built, tested and deployed the recommendation system that decides whom to contact, when and on which channel; conversion up by as much as 30% across 3 million customers and 50 million interactions.
    • Data pipelines, evaluation methods and monitoring for the campaigns and the models.
    • The MLOps stack behind it, GitLab CI/CD, Docker on AWS EC2, Kubernetes, Airflow and MLflow.
    • +30% relative conversion uplift telecom client, engine vs. previous targeting
    • +5–15% monthly conversion uplift bank client, drip campaigns on 500–1,000 leads a day, engine on vs. CRM targeting
    • 3M customers scored 50M interactions processed in production
    PythonCausalMLXGBoostAirflowMLflowDockerKubernetesAWSGitLab CI/CD
  5. May 2023 – Oct 2023 Paris, France

    Data Scientist, Geo and NLP team

    namR · Data-intelligence startup, buildings and energy transition

    • New features on datasets of up to 50 million rows on Google Cloud.
    • Cut the solar-panel detection pipeline from 25 hours to 8 by replacing its full-factorial combination plan with an optimised Latin-hypercube design.
    • 25h → 8h solar-panel detection runtime full-factorial plan replaced by an optimised Latin-hypercube design, same input data
    PythonSQLGCPBigQueryPySpark
  6. Oct 2019 – Dec 2022 Palaiseau, France

    Research Engineer, AI and Data Science R&D (industrial PhD, CIFRE)

    TotalEnergies One Tech · R&D branch of TotalEnergies

    • Gaussian-process models to forecast gas-well production, with calibrated 80% prediction intervals.
    • Causal-inference models for geothermal wells, which led to a patent application (US 2024/0211803 A1) and three publications.
    • Regular presentations of the results to the Digital Factory stakeholders.
    PythonRMATLABGaussian processesCausal inferenceAzure

Education

  1. 2019 – 2022

    PhD in Applied Mathematics

    École Polytechnique, Institut Polytechnique de Paris

    Causal inference and uncertainty quantification. Industrial PhD (CIFRE) with TotalEnergies. Thesis "Statistical learning and causal inference for energy production", defended November 2022.

  2. 2016 – 2019

    Engineering degree (diplôme d'ingénieur)

    École des Mines de Saint-Étienne

    Major in Data Science, minor in Operations Research and High Performance Computing.

  3. 2018 – 2019

    MSc in Applied Mathematics (MMMEF)

    Université Paris 1 Panthéon-Sorbonne

    Mathematical Models and Methods in Economics and Finance. Stochastic calculus, optimisation, credit risk.

Skills

Methods
Causal inferenceUplift modellingA/B testing and experiment designRecommender systemsBayesian modellingUncertainty quantificationGaussian processesOptimisationTime series
Languages
PythonRSQLMATLAB
Libraries
scikit-learnXGBoostLightGBMCausalMLpandasNumPyPySparkSQLAlchemyMatplotlibggplot2
MLOps and production
DockerKubernetesAirflowMLflowDVCGitLab CI/CDFastAPIStreamlit
Cloud and data
AWS (EC2RedshiftS3)GCP (BigQueryVertex AI)AzurePostgreSQLMySQLTrinoDataiku DSS
Tools
JupyterVS CodePyCharmDataiku DSSGrafanaJiraConfluence

Summer and winter schools

  1. Aug 2021
    Machine Learning Summer School (MLSS) 2021

    National Taiwan University, Taipei (online)

  2. Jul 2021
    Hi! PARIS Summer School on AI and Data for Science, Business and Society

    Institut Polytechnique de Paris and HEC Paris

  3. Sep 2020
    Gaussian Process and Uncertainty Quantification Summer School

    The University of Sheffield

  4. Jan 2020
    Data Science, Optimization and Operations Research Winter School

    École Polytechnique Fédérale de Lausanne

Awards and scholarships

  1. Feb 2018
    Winner, IMT-Total data science challenge

    Kaggle-style competition won with random-forest models; awarded a research internship with TotalEnergies.

  2. 2020-2022
    Study Abroad Grant (3,100€)

    Moroccan Ministry of Higher Education.

  3. 2016-2019
    Government Merit Grant (15,500€)

    Moroccan Ministry of Higher Education.

Languages

Arabic
native
French
bilingual
English
proficient
Spanish
elementary