Rift Valley Fever (among other zoonosis) have been worth to monitor from different bodies (Research, Governments, WHO, FAO, WOAH, etc.). However, traditional monitoring system based on the public health system lacks efficiency for different reasons (amount of data collected and their real-time nature). Several solutions based on AI/ML and Deep Learning (DL) have been proposed recently (sample processing, detection of environmental factors, vector monitoring, etc.). However, providing large data collections (in real time if possible) and reliable for these DL based solutions is a big challenge. In this paper we address and propose to monitor RVF vectors (Aedes, Anopheles and Culex) using an Edge-AI Framework for Citizen science. It mainly consists of a mobile application deployed on the smartphone of breeders, citizens, volunteers and clinicians. The mobile application embeds a customized DL model for RVF vector classification. The main outcomes of this paper are: i) design of an Edge-AI framework for data collection and vector monitoring; ii) a deep learning model for mosquito species classification to facilitate citizen science and iii) a mobile application embedding the ML Model. Efficientnet0, MobileNetV2 and Resnet50 has been tested for vector classification and we achieved an accuracy of 95.56% with a tuned and modified version of MobileNetV2 with the benefit of being mobile app friendly with a lighter size and faster.