The security situation in Kenya has deteriorated over time due to the low number of police personnel in the country which is currently at a population ratio of 1:1150. Security challenges have increased from mere theft to carjacking attacks and to more serious and evolved challenges like murder and terrorism. The government’s efforts towards reducing these crimes have been ineffective as there are no mechanisms for gathering intelligence at low levels. Intelligence gathering especially from the public is very essential in tackling matters to do with insecurity. This research proposes a simple, convenient and efficient solution to the security challenges that Kenya is currently facing with respect to systematic gathering of intelligence and its analysis by the use of a mobile and web based application. The mobile based solution will integrate the use of GPS location services and ensure that it uses machine learning by using predictive models produced from Multiclass Decision Forest Algorithm, and to be able to provide detailed descriptive statistical analysis, text mining analysis of criminal activity taking place, as well as prediction analysis to predict crime patterns. The solution has an administrative web-based backend that will be accessed by the police force to ensure they get detailed information of criminal activities. From this portal, tests were done by entering information regarding suspicious person, potential suspicious person names associated with the submitted information are provided together with relevant scores to depict the most likely accurate name.