It is very inconvenient to take a blood sample from a diabetic patient’s fingertip to assess glucose. Blood-related infections are more likely to occur when invasive methods like one-touch glucometers and laboratory tests are used. In the current paper, we design a unique non-invasive blood glucose measurement as a potential solution to this significant problem. Hemoglobin,a protein in red blood cells (RBCs), transports oxygen throughout the body. A person’s hemoglobin gets glycosylated in response to increased blood sugar. Glycated hemoglobin (HbA1c) remains in our blood for around 90 d (three months). This study detect glucose molecules in human blood using a near-infrared (NIR) spectroscopy approach at 940 nm wavelengths. The Near Infra Red spectroscopy and many highly accurate machine learning (ML) models are the foundation for this gadget. A Linear Regression model has been presented for precise measurement. The calculated HbA1c values have been compared with HbA1c values from the Apex Hospital Laboratory, Jaipur. The mean absolute relative difference (MARD) and average error(AvgE) concerning the predicted hbA1c values have been reported to be 17.32% and 19.85%, respectively. The suggested non-invasive spectroscopic apparatus offers a precise and economical solution.