You're seeing this page as if you were . The main menu is still yours, though. Exit from immersion
Hedi MasmoudiHM

Hedi Masmoudi

Data Scientist | Generative AI expert

€300/day
Paris, FR
3-7 years

Average response time: 1 hour

Freelancer profile translated to English.
Back to original language

About Hedi

I am a consultant specializing inDataandArtificial Intelligence**, with particular **expertisein the design anddeployment of generative AI solutions**. I work on the **entire**project lifecycle**, from scoping to production, using agile methodologies (**Scrum**) and technologies such asMachine Learning**, **LLMs**, **RAG systems**, and **Deep Learning**. Passionate about **applied**mathematics** and **statistical models**, I have a strong appetite for advanced data analysis and the creation of AI solutions with high business value.
  • English

    Native or bilingual

  • French

    Native or bilingual

Can work on-site
Paris (up to 50km)

Experience

  • Talan
    Data Engineer / Developer
    CONSULTING AND AUDITS
    March 2025 - October 2025 (7 months)
    Paris, France
    As part of a strategic internal project at Talan, I led and executed from end-to-end an intelligent staffing chatbot capable of identifying the most relevant consultants for a given mission offer. I implemented a RAG (Retrieval Augmented Generation) engine deployed on Microsoft Teams, connected to a vector index hosted on Azure.

    • Designed a completeCV processing pipeline**: **ingestion**, **cleaning**, format **harmonization**, automated **metadata extraction(availability date, grade, current mission, etc.).
    • Deployed astorage architectureon **Azure Data Lake**, with software structuring of files (original CVs, anonymized CVs, mappings, dynamic statuses, etc.).
    *Automated CV anonymization(name removal from SharePoint hierarchy) in compliance withGDPRrequirements.

    *CV vectorizationusing OpenAI embeddings (text-embedding-3-large) and creation of an enriched FAISS index.
    • Implemented aRAG(Retrieval Augmented Generation) search engine withavailabilityfiltering, **vector similarity**, **contextual sorting**, and **custom scoring**.
    • Implemented anintent classification agentfor targeted questions about selected consultants (**experience**, **availability**, **mission fit**).
    *Containerizedthe project usingDockerto ensure chatbot portability and scalability.

    *Deployedthe chatbot onMicrosoft TeamsviaAzure Bot Service(application registration, channels, secrets configuration).
    Microsoft Azure RAG OpenAI Langchain LLM
  • AXA
    Gen AI Developer / Front End
    June 2024 - February 2025 (8 months)
    Paris, France
    Within the AMMRAG project (Agent Multi Modal RAG), the objective was to create achatbotspecialized in a domain (**Insurance**) capable of answering questions precisely and without error on various document types (**complex tabular data**, **Word**, **PDF**, **scanned PDF**, **images**, **PPT**).

    Development and improvement of FAISS (Vector Store)
    • Enriched FAISS by indexing new documents to improve information retrieval.
    • Processed and ingested various document formats (PDF, Excel, OCR, Word).
    • Implemented an automated FAISS update process, ensuring efficient indexing of new data.

    Automation of tests to validate response relevance
    • Developed a test automation script to evaluate the quality of the model's responses.
    • Compared generated responses with expected answers, facilitating the evaluation of RAG result relevance.
    • Integrated a scoring mechanism and analysis of RAG response performance.

    Development of a graphical interface for consultation and interaction
    • Designed and developed a graphical interface allowing users to test and interact with the model.
    • Integrated several features: document type selection, button to display documents used for the last RAG response, button to download conversation history.
    • Implemented a fluid and intuitive user experience to simplify access to documents and their content.
  • Nestle
    Data Sciences / ML/MLOps
    January 2024 - March 2024 (2 months)
    Paris, France
    As part of a collaboration withNestléin 2023, the project aimed to improve and supportMachine Learning pipelinesviaAzure Machine Learning**. The objective was to **optimizetheprocessesof datapreprocessingand **model training**, while ensuring better monitoring and maintenance of deployed solutions.

    Optimization and monitoring of Machine Learning pipelines
    • Developed detaileduser storiesto define **user needs**, thus facilitating feature prioritization and alignment with Nestlé's strategic objectives.
    • Participated inPoker Sessionswith the team to estimate task complexity and refine priorities, ensuring efficientbacklogmanagement and timely delivery.
    *Optimizedthecleaning**, **transformation**, and **preparation**processes** fordatato makemodel trainingmore efficient and automated on the Azure Machine Learning platform.
    • Implementedmonitoring strategiestotrackmodelperformancein production and ensure **proactive maintenance**.

Recommendations

Be the first to recommend Hedi

Help this freelancer shine by sharing your experience working together.

These freelancer profiles also match your criteria

AgathaA

Agatha Frydrych

Backend Java Software Engineer

4.7

(3)

2

BaptisteB

Baptiste Duhen

Fullstack developer

4.6

(4)

5

AmedA

Amed Hamou

Senior Lead Developer

4

(2)

7

AudreyA

Audrey Champion

Web developer

4.3

(3)

4

Education

  • Master in Data Sciences & Business Analytics
    Centrale Supélec X ESSEC BUSINESS SCHOOL
    Master en Data Sciences & Business Analytics
  • Bachelor in Political Economy
    HEC Lausanne
    Bachelor en Economie Politique

Skill set

Categories