The demand for global energy will increase by 1.5 times the current demand in the next 30 years primarily driven by population and economic growth, particularly in developing Asian and African nations. The world energy mix consists of 75% fossil fuels. While fossil fuels remain the primary source, reducing their carbon footprint is the focus going forward. Efficient use of technology to reduce greenhouse gas emissions is the primary goal for researchers in the energy industry. While progressing on that front, we have higher than-normal levels of greenhouse gases already present in the atmosphere due to past actions. Carbon capture, utilization and storage (CCUS) techniques have been invented in recent years to remove carbon from the atmosphere and use it for alternate purposes. Using CO2 for new products (plastics, fuel, etc.) may circulate carbon in the environment. The most efficient and effective way to remove carbon from circulation is geological storage. The petroleum industry has valuable expertise and equipment to safely sequester CO2 underground and monitor it effectively. CO2 storage with enhanced oil recovery (EOR) provides a reliable solution and is a commercially viable process. Numerical simulation models are developed to understand the sub-surface behavior of fluids. Machine learning (ML) and artificial intelligence (AI) advancements have boosted numerical simulation performance and run time. Using hybrid proxy ML models, the simulation time is reduced by orders of magnitudes enabling faster implementation and turn-around time. This paper uses various ML techniques along with a traditional reservoir simulation approach to estimate CO2 storage potential with incremental original oil-in-place (OOIP) of an Indian stacked pay oil reservoir.