Health and safety considerations of indoor occupants in enclosed spaces are crucial for building management which involves the strict control and monitoring of carbon dioxide levels to maintain acceptable air quality standards. For this study, we developed a wireless, noninvasive, and portable platform for the continuous monitoring of carbon dioxide concentration in enclosed environments, i.e., academic rooms. The system aimed to monitor and detect carbon dioxide using novel low-cost metal oxide-based chemoresistive sensors, achieving sensing performance comparable to those of commercially available detectors based on optical working principle, e.g., nondispersive infrared sensors. In particular, a predictive study of carbon dioxide levels was performed by exploiting random forest and curve fitting algorithms on chemoresistive sensor data collected in an academic room, then comparing the results with lab-based measurements. The performance of the models was evaluated with real environment conditions during 7 weeks. The field measurements were conducted to validate and support the development of the system for real-time monitoring and alerting in the presence of relevant concentrations (above 1,000 ppm). Therefore, the study highlighted that the curve fitting model obtained was able to recognize with an