Freelancer profile translated to English.
Back to original languageAbout Benbacer
I am a Data Scientist and AI Engineer with over 6.5 years of experience in designing, developing, and industrializing artificial intelligence, machine learning, and generative AI solutions.
I help companies transform their data into concrete performance levers, whether it's to optimize their processes, better understand their users, automate business tasks, or integrate advanced AI capabilities into their internal applications.
My added value lies in a dual expertise in Data Science and MLOps/LLMOps, which allows me to be involved in the entire project lifecycle: understanding needs, data exploration, modeling, model development, production deployment via APIs or CI/CD pipelines, monitoring, and continuous improvement. I have already supported numerous projects involving forecasting, scoring, user segmentation, anomaly detection, NLP, embeddings, OCR, behavioral analysis, recommendation engines, and integration of RAG pipelines with LLMs.
I place particular importance on simplification and collaboration with non-technical teams to ensure that each solution truly meets a business need and delivers measurable value.
If you are looking for a profile capable of quickly understanding your challenges, proposing robust and innovative approaches, and delivering ready-to-use, reliable, and results-oriented AI/ML solutions, I would be delighted to assist you.
Arabic
Native or bilingual
French
Fluent
English
Fluent
Can work on-site
Paris (up to 50km)
Experience
- Saint Gobain Distribution Bâtiment FranceData ScientistRETAIL (SMALL BUSINESS)June 2022 - Today (4 years)Paris, France
- Design and deployment of predictive and simulation models to optimize pricing strategies and demand forecasting in multi-segment and multi-seasonal environments.
- Performing statistical modeling to quantify elasticity and exogenous impacts (seasonality, price, regional variations), providing actionable KPIs for strategic decision-making.
- Development and automation of Machine Learning models (GBM, LSTM, Random Forest, KMeans) with Python, AzureML, and Cloudera, enabling the scaling of customer recommendations from a few thousand to tens of thousands per day with over 96% accuracy.
- Creation of data pipelines and automated dashboards (Python, SQL, MicroStrategy, SAP) ensuring the reliability and consistency of pricing and marketing analyses.
- Redesign of a struggling recommendation algorithm through mathematical reformulation and model redesign, generating a measurable impact on sales.
- Development of an OCR solution based on a pre-trained model to automatically detect and correct the orientation of product images, improving data quality and the accuracy of dimensions displayed on the website.
- Adjustment and deployment of a pre-trained NLP model to automatically extract product features from unstructured PDF catalogs, reducing manual data entry time by 80%.
- Experimentation with a GPT-type model (LLM) for automatic product categorization and taxonomy validation, integrating prompt engineering techniques to test semantic similarity and classification robustness.
- Management of model industrialization: Python, Spark, Databricks, CI/CD, MLflow, GitLab CI, Azure DevOps, with performance and model drift monitoring.
- Training business teams in Machine Learning, strengthening data culture.
- StradaDATA ScientistTECHApril 2019 - May 2022 (3 years and 1 month)Bressuire, France
- Design of probabilistic and time series models for IoT sensor reliability and predictive maintenance of transport fleets (time-weighted Naive Bayes, Markov models…).
- Design and deployment of an NLP chatbot for after-sales service with Rasa and Dialogflow, using Word2Vec and TF-IDF to detect recurring patterns in customer tickets and propose automated resolutions.
- Performing statistical analyses and A/B testing to validate hypotheses and optimize customer interaction workflows.
- Design and deployment of ML pipelines (LSTM, GBM, Random Forest) with MLflow and Docker, integration of REST APIs for operational use.
- Creation of decision-making dashboards in React + Toucan Toco, connected to ML pipelines via APIs, enabling data storytelling and customer reporting focused on decisions.
- Development of ARIMA, LSTM, and LightGBM models for forecasting mileage and driver performance.
- Design of a hybrid KNN algorithm with statistical outlier filtering, improving robustness on imbalanced datasets.
- Implementation of ETL workflows in Python (Pandas, SQLAlchemy) to prepare and standardize transport data.
- Creation of interactive dashboards (Plotly, Dash, Mapbox) for monitoring and analyzing fleet performance.
- Performing exploratory and predictive analyses on vehicle telematics data to identify and predict abnormal driving behaviors.
- Design of a supervised ML pipeline (Random Forest, XGBoost) to predict delivery delays.
- Creation of visual indicators for internal reporting and product demonstrations.
Recommendations
Be the first to recommend Benbacer
Help this freelancer shine by sharing your experience working together.
These freelancer profiles also match your criteria
Agatha Frydrych
Backend Java Software Engineer
4.7
(3)
2
Baptiste Duhen
Fullstack developer
4.6
(4)
5
Amed Hamou
Senior Lead Developer
4
(2)
7
Audrey Champion
Web developer
4.3
(3)
4
Education
- Master 2 in Data ScienceUniversité Aix Marseille2020- Logiciels et langages de programmation : Python, R - Analyse des données, visualisation des données, traitement des données, apprentissage automatique, deep learning, traitement de signal - Statistiques, Optimisation mathématique, Statistique Géospatiale
- Master 2 in Mathematical and Actuarial Statistics EngineeringUniversité Aix Marseille2019- Logiciels et langages de programmation : R, VBA, SQL, SAS - Gestion des risques, IARD, actuariat, gestion de portefeuille - Statistiques, Analyse des données, apprentissage automatique, séries temporelles