Antoine Dejoie

Used new LLM such as Gemini to construct optimized pipelines to process business and sales data in real time. Conducted ML analysis for forecasting and anomaly detection.

Development of digital products for satellite image analysis within the R&D team, thanks to the implementation of new neural network architectures (fusion of CNN and RNN).

My main task was to use artificial intelligence methods to improve the insider threat detection system. This system uses information reported in real time by thousands of sensors placed in the customer's computer network, to detect the possibility of dangerous, even malicious, activity within the organization. My secondary mission was to help out with the company's other research projects. I joined the R&D team, which translates scientific concepts into practical methods, and worked closely with the various engineering teams, who integrate these methods into customer solutions.

Worked in an MIT start up as a software engineer to develop new tools of data analysis using AWS and machine learning to improve yields and detect leaks.

My task was to help optimize the process for handling the Finance Department's proprietary financial transactions (swaps, securities, holdings), covering the entire front office, middle office, back office and custodian chain. During my 4-month internship, I was entrusted with several modules: process and procedure description, change management (user acceptance and optimization of the existing workflow tool), development of local tools: front-office dashboard, automation of back-office and accounting tasks, etc.