This dissertation explores the advancement of exoplanet spectroscopy, focusing on methods to improve signal-to-noise ratio (SNR) and maximize the scientific yield of observations. With over 5,700 confirmed exoplanets, ranging from hot Jupiters to Earth-like planets, spectroscopy is crucial for analyzing their atmospheres, temperatures, and potential habitability. I begin with an analysis of the hot Jupiter WASP-79b, where systematic errors in spectroscopic data from Hubble’s Wide Field Camera 3 were carefully treated. Results from the analysis show a planet with a brightness temperature of about 1900 K. Combining the data with Spitzer measurements and comparing it to atmospheric models, indicates a variety of potential atmospheric scenarios, including equilibrium chemistry with solar metallicity and non-equilibrium models featuring high-temperature clouds or significant H- opacity. I then explore tools developed to improve the science return for the upcoming NASA mission, Pandora designed to study exoplanet atmospheres and the impact of stellar contamination using transmission spectroscopy. The first is the development and integration of a 2-D spectrum simulator and pseudo-retrieval framework for the near-infrared channel of Pandora, which provides a quantifiable approach to refine target selection and optimize observation strategies. Second, a metaheuristic-based scheduling algorithm was developed to ensure efficient use of Pandora’s observing time, balancing various factors such as observing efficiency and contamination by the South Atlantic Anomaly. Preliminary testing of Pandora’s near-infrared detector has also yielded important reference data for future observations, including initial characterization of some noise sources we can expect to see and can start developing treatment procedures for ahead of launch. This research also investigates 1/f-noise mitigation techniques for JWST’s NIRSpec BOTS observations. Using the NSClean package, the study found that applying the Fourier-space noise treatment approach at the integration level data products yielded the greatest reduction in noise. This study also found that the choice in masking routine used to identify the background region needed to characterize the 1/f-noise played a critical role in the treatment process. Together, these contributions advance the understanding of exoplanet atmospheres and improve observational strategies for future missions.