Millian Lamiaux

Fine-tuning an AI model to detect secrets in source code. Developing a desktop application to embed the model on employees' laptops. Working on the security of LLMs (Large Language Models) and the protection of sensitive data. Collaborating with teams to reinforce security best practices in AI projects.

This project focuses on the design and deployment of a chatbot based on the RAG (Retrieval-Augmented Generation) architecture, a cutting-edge technology that links the power of language models to an up-to-date knowledge base. The uniqueness of this tool lies in its ability to analyze phishing emails and SMS submitted by users in real-time. By cross-referencing these threats with cybersecurity databases, the chatbot provides precise and contextualized responses, transforming each attack attempt into an interactive learning opportunity for employees.

Using Llama 2 and GPT-3.5 to automate the generation of standardized content. Optimizing product descriptions for the company's webshop. Improving the efficiency and consistency of marketing content.