Nagbe Komi

I develop Artificial Intelligence solutions for companies. We have developed an AI platform that allows companies to visualize data easily and intuitively, and to manage data quickly and efficiently. Our platform integrates solutions such as AI-assisted code generation and survey and poll generation for market research. To enable companies in the energy sector to better predict their energy portfolio in terms of energy production and consumption, we have integrated an advanced energy forecasting system into our platform that uses the most advanced artificial intelligence technologies. Our platform is a SaaS.

Gas and electricity forecasting activities are at the core of the success of Eni's main business, which is energy supply. Gas and electricity forecasting was outsourced outside Eni France before our arrival in the company. We were therefore recruited to enable the internalization of this activity. Part of our mission was to create machine learning models to predict gas and electricity consumption across all customer segments. To achieve this mission, we built the customer portfolio, estimated customer churn rates, and optimized the volume to be injected into the energy grid based on various parameters. In addition to forecasting activities, we participated in the restructuring of Eni's information system in terms of structuring databases under PostgreSQL, identifying relevant information to collect, and how to use the collected data. We were involved in all strategic discussions on new laws and regulations in the energy sector and the future evolution of the electrical grid. The main technologies used are R, Python, PostgreSQL, PgAdmin, MySQL, GitLab.

This research work was done for the company Rouge Gorge. The mission consisted of segmenting the purchase data and customer data of this company. Purchase data includes receipts identifying the different items purchased by the company's customers. Customer data defines the journey of each customer over several years of visiting the company's stores. We were faced with large volumes of data. To succeed in our mission, we used certain machine learning techniques and significant computing power.