The data cleaning process is a fundamental step in Data Science and Big Data analytics because raw data collected from real-world sources is often incomplete, inconsistent, noisy, or duplicated. Before applying machine learning algorithms or statistical analysis, data must be transformed into a reliable and structured format. In large-scale environments such as those used by organizations like Google and IBM, data cleaning becomes even more critical due to the volume, velocity, and variety of data. Poor quality data can lead to incorrect insights, biased predictions, and faulty decision-making. Data cleaning typically begins with data understanding and profiling. In this stage, data scientists examine dataset characteristics such as data types, distribution patterns, outliers, and missing values. Tools and distributed frameworks developed by communities like the Apache Software Foundation help process massive datasets efficiently. Profiling helps identify errors such as inconsistent date formats, invalid entries, or unexpected null values. Without this initial assessment, cleaning efforts may miss hidden data issues that affect downstream analysis. The next step involves handling missing values, which are one of the most common problems in real-world datasets. Missing data can occur due to system errors, manual entry mistakes, sensor failures, or incomplete surveys. The choice of imputation technique depends on the data type, missing pattern, and business context. One simple method is deletion, where records with missing values are removed. While easy to implement, deletion may lead to loss of valuable information and biased results if missing data is not random. For example, removing patient records with missing medical measurements might distort healthcare model predictions. Mean, median, and mode imputation are basic statistical techniques. Mean imputation