Electroencephalography (EEG) has been recognized as one the finest cost-effective techniques to measure the electrical activity of the human brain. Since the electrical activity measured at the scalp is an abstract representation of the regional brain activity of the human brain, the EEG readings (EEG data) are often useful in many aspects. The value of an EEG data is proportionate to an electrical potential measured at a specific position on the scalp using an electrode. One of the major uses of EEG data from the medical domain is detecting the brain disorders such as Epileptic Seizure. Brain Computer Interfacing (BCI) is another use of EEG data. The paper discusses the analytical procedures of EEG data acquired using consumer-grade low-cost EEG devices in order to achieve the objective of implementation of a cost-effective BCI system. Furthermore, the paper presents a critical evaluation of some selected consumer-grade EEG devices from the point of view of cost-effectiveness against the expected accuracy. We have obtained a maximum of 76.6% for user intention detection with a low-cost EEG device Emotiv's “Insight”. This is a remarkable achievement in a context that, the maximum accuracy reported for the EEG devices in the class was only 60.57%. We emphasize the key impact behind this achievement is the employment of the technique Multi-Agent Systems (MAS), which is an Artificial Intelligence (AI) approach, for EEG data analysis and intention classification.