Understanding the dynamics of inflation is important for the policymakers, in order to employ effective monetary policy that can keep the economy healthy. This paper employed two models, i.e., Random Forest (RF) for precision in inflation prediction and Generalized Additive Models (GAM) for understanding the determinants of inflation. GAM captures nonlinear associations among variables, whereas RF is a robust machine learning algorithm which supports improved prediction. While studying month-wise data on macroeconomic trends over the past two decades in India, the significant drivers of inflation were the money supply and exchange rates. The effects of changes in the repo rate rely on the existing economic environment. Crude oil prices have little direct impact. This paper shows how important it is to use a variety of analytical techniques in order to get accurate economic data and inflation forecasts that can be understood. This is necessary for maintaining economic stability and smart decision-making on the monetary policy.