Growing interest in Indoor Air Quality (IAQ) has attracted considerable attention since the COVID19 pandemic. To maintain a good IAQ, both gaseous and particulate contaminants should be maintained below an acceptable limit in an indoor environment. IAQ is not easy to predict, given that most indoor air gas pollutants are present simultaneously within an indoor built environment at variable concentrations. Therefore, it is difficult to assess or classify these mysterious gaseous pollutant concentrations and hazard levels. This paper presents the design of an accurate IAQ index standard based on acceptable limits for the level of multiple indoor air pollutants when designing Heating, Ventilation and Air Conditioning (HVAC) systems. These limits will be used in the deep learning process in the Artificial Neural Network Fuzzy Inference System (ANFIS). Cognitive expertise is used to create the rules of fuzzy inference that generate the final risk level. The ANFIS-based controller uses real-time data obtained by IAQ sensors to generate the final value of the resulting hazard level. The resulting hazard level indicator can be fed to a microcontroller for monitoring, alarms, and control purposes. The microcontroller can generate an operating signal for the HVAC system to maintain an acceptable indoor air level. As a result, a safe IAQ level is maintained, which will reduce the cost of the low throughput of poor ventilation systems.