WQI (Water Quality Index), a standard indicator to appraise and describe quality of surface water. This research offers insights into the groundwater quality and concentrations of heavy metals in Gurugram city groundwater. The physical and chemical parameters such as pH, TDS, EC, \(Na^+\) , \(Ca^{2+}\) , \(K^+\) , \(Mg^{2+}\) , Cl \(^-\) , \(HCO_3^{2-}\) , \(SO_4^{2-}\) and heavy metals such as Cd, Zn, Fe, Pb, Cr, Ni, Cu were considered for computing WQI. Concentrations of heavy metals in groundwater assessed utilising an AAS (Atomic Absorption Spectrophotometer). This research introduces utilisation of ANN (Artificial Neural Network) and MLR (Multiple Linear Regression) as predictive models for estimating WQI values in Gurugram, Haryana, India. The outcomes were achieved through the subsequent data partitioning: 70% for training, 15% for validation and another 15% for testing, facilitated by Neural Network Fitting application. Analytical results confirmed that none of the samples were considered to be in the categories of excellent water and good water. This is mainly because of the introduction of pollutants from domestic or agricultural sources. The groundwater’s WQI (Water Quality Index) suggests that it is unsuitable for drinking purpose because of microbial contamination, excessive nutrients, salinity, industrial activities. The model’s performance was assessed using qualitative and quantitative indices, specifically the Regression Coefficient and RMSE (Root Mean Square Error). The most favourable modelling outcomes were achieved using a network comprising 5 neurons in hidden layer. The coefficient of determination, R \(^2\) = 0.93352794, signifies a highly favourable alignment with the data model and an elevated correlation coefficient of 0.987, evident both in overall assessment and across individual subsets, along with minimal MSE values (training, test and validation) and an RMSE of 0.002966 attained.