fr
en
Advanced Data Modelling: From Analysis to an AI Project
Your analyses need structured data and suitable models. Connect exploration, modelling and evaluation to turn a business question into an analytical approach. Strengthen your ability to explain results and the limitations of the methods used.
- Duration
- 3 days 21 hours
- Code
- DATA006FR Code
Presentation
At a time when AI generates more data than it analyses, knowing how to turn a raw file into a reliable model has become a critical competitive advantage. This intensive 3-day course combines supervised and unsupervised modelling with the art of prompting to accelerate data cleaning, feature engineering and interpretation of results.
Working in pairs, you will cover the complete modelling project lifecycle: importing data, quality auditing, training, validation, packaging and deployment. Each stage is illustrated by a real case—employee departure prediction, fraud scoring or customer segmentation—and a Python/Excel template you can reuse when you return to work.
By the end of the course, you will have a completed project, an ethics checklist aligned with GDPR and the AI Act, and a roadmap for taking your model into production. These resources support rapid, responsible AI adoption within your organisation.
Objectives
- identify the business challenges and opportunities associated with data modelling;
- use common analytics platforms and modelling libraries;
- select, train and evaluate supervised and unsupervised algorithms suited to your data;
- accelerate data cleaning, enrichment and documentation with generative AI tools;
- respect ethical and regulatory principles throughout the data lifecycle;
- plan, execute and present a complete modelling mini-project, from scoping to communicating results.
Program
Module 1: preparing data with generative AI
- Data modelling fundamentals.
- Distinguishing conventional AI from generative AI.
- The 5 Vs of Big Data and their implications.
- Overview of generative AI tools: LLMs, GPT, Claude, Gemini and others.
Demo
- Automated dataset exploration using GPT.
Practical exercises
- Explore and visualise a dataset with Python and GPT.
Module 2: cleaning and exploring data
- Key data preparation steps: cleaning, handling missing values and encoding.
- Using generative AI to create cleaning scripts and suggest visualisations.
Practical exercises
- Clean a dataset and prepare it for modelling.
Module 3: building, evaluating and interpreting models
- Introduction to modelling methods such as regression and random forests.
- Model evaluation techniques: cross-validation, performance metrics and ROC curves.
- Using generative AI to create modelling code and interpret results.
Practical exercises
- Build a model and interpret its results with AI assistance.
Module 4: managing scalability and large datasets
- The architecture and logic of distributed platforms such as Spark.
- AI use cases for managing large datasets.
- Using generative AI to create PySpark scripts, optimise pipelines and summarise transformations.
Practical exercises
- Create a modelling pipeline for a large dataset.
Module 5: establishing governance and ethics
- Key regulations such as GDPR and ethical issues such as bias and auditability.
- Introduction to explainability tools: SHAP and LIME.
- Using generative AI to produce documentation, ethics notes and internal charters.
Practical exercises
- Audit an AI model with an explanatory report generated by GPT.
Module 6: completing a full modelling project
- Key project stages: structuring, preparation, modelling and documentation.
- Forming teams for a practical project, such as employee departure prediction or customer segmentation.
Practical exercises
- Structure and carry out a modelling project from start to finish.
- Submit the final deliverable: an augmented analysis portfolio including code, visualisation and a report supported by generative AI.
Audience
This course is intended for professionals seeking to develop data modelling and AI skills, including:
- Executives and business managers who want to understand data-related challenges.
- Data or digital managers seeking technical expertise.
- Analysts and research officers who want to deepen their skills.
- IS or innovation project managers who need to master AI project lifecycles.
- Professionals retraining for data roles.
Prerequisites
This course requires the following prerequisites:
- Data handling skills: proficiency in Excel or an equivalent spreadsheet application, and the ability to read activity reports and build or interpret KPIs.
- Basic statistics: understanding of correlation, measures of central tendency such as mean and median, and the principle of simple regression.
Teaching and assessment methods
- Initial skills assessment
- Training materials provided to participants
- Continuous assessment throughout the course
- End-of-course feedback questionnaire
- Combination of theory and practical application
- Attendance records
- Post-course follow-up evaluation
- Practical exercises
Course highlights
- Operational approach: the course focuses on integrating generative AI to optimise and automate modelling tasks.
- Advanced tools: use popular tools and languages such as Python, PySpark and Databricks, along with high-performing generative AI models such as GPT-4 and Claude.
- Comprehensive programme: cover the complete data lifecycle, from ethical preparation to delivery of a full project.
- A practical project: conclude with a collaborative project that lets you immediately apply all acquired skills in a professional context.
Dates and sessions
Choose the date and delivery format that suit you.
No upcoming sessions are currently available.
GPT is a registered trademark of OpenAI, Inc.
Claude is a registered trademark of Anthropic PBC.
Copilot is a registered trademark of Microsoft Corporation (French website).
Gemini is a registered trademark of Google LLC.
We are neither endorsed by nor affiliated with any of these companies. The tools are mentioned as examples and for illustration.