It is paramount that solutions to questions of energy security for a developing nation be addressed through its internal resources. India, endowed with 23.8 billion tons of deep un-minable lignite, faces the challenge of economically sustainable extraction. This study presents a comprehensive assessment of lignite's suitability for underground coal gasification (UCG) compared to bituminous coal. Employing a multi-dimensional approach, combining single-step and distributed activation energy model of pyrolysis kinetic modeling with extensive physicochemical analysis (proximate and ultimate analyses, FTIR, SEM–EDX, XRD), revealed that lignite has a lower activation energy making it suitable for UCG. This finding, highlighted by kinetic modeling, is substantiated by the lignite’s structural properties as identified in physicochemical analysis. This study leverages machine learning for higher heating value prediction, finding long short-term memory as the most effective model compared to five other models based on the R² score and error values. Additionally, an XGBoost algorithm-based model predicts syngas heating value and yield while showcasing the application of machine learning in enhancing energy prediction accuracy. The economic analysis, applied for a 50 MW power plant framework, determines the unit costs for syngas and electricity production to be 7.49 and 6.71 $/GJ and 53.68 and 59.93 $/MWh for the samples of lignite and bituminous coal, respectively. The sensitivity analysis revealed that the energy content in syngas is the most significant parameter. These comprehensive findings validate lignite's potential for energy production in India, offering insights for similar resource optimization in other developing countries.