Vital sign monitoring is the process of recording human physiological signals in order to determine the mental stress level. High stress levels can prove to be dangerous especially for certain individuals such as autistic children who are not able to express mounting levels of stress before it leads to a full anxiety attack. This paper presents the prototype design of a real-time embedded device that accurately measures heart rate and galvanic skin response (GSR) in a non-invasive and non-intrusive way which is then used by the intelligent decision making module that uses fuzzy logic to determine the stress level of the user. Such a device could be used with autistic children in order to give early warning of an impending anxiety attack and help adults to prevent it from happening. The prototype was designed using Arduino mega platform and tested with 35 clinical patients in three experimental settings targeted to induce low stress, medium stress and high stress response. Initial results have shown that the device is capable of detecting and displaying the various stress levels efficiently.