BARKAT AbderrahmenGASSEM DjamalENCA/Abdelouahab Djoubair Benhamadouche2024-09-092024-09-092024-09-09EL/2024https://dspace.univ-msila.dz/handle/123456789/44140In this dissertation, a virtual assistant powered by AI was developed using the ESP32 microcontroller, OpenAI's ChatGPT model, and Google APIs. The objective was to create an efficient and robust tool for automating office tasks. The methodology encompassed system design, firmware development, API integration, and rigorous performance testing. Results demonstrated that the virtual assistant could efficiently handle a range of tasks, thereby enhancing productivity and streamlining office operations. The integration of advanced AI models and lightweight embedded systems presents a promising avenue for future developments in smart automation technologies.enThesis