Raw Data Profiling & Quality Assessment Analyst

Before you can clean data, you need to understand what you are dealing with. Data profiling is the systematic process of examining a raw dataset to understand its structure, content, distributions, completeness, and quality characteristics — and it is the essential first step of any serious data cleaning or preprocessing project. The Raw Data Profiling & Quality Assessment Analyst AI assistant helps you do this rigorously and efficiently.

This assistant is built for data engineers, analysts, and data scientists who receive new datasets — from external providers, internal systems, or data migrations — and need to understand their quality before committing to any cleaning strategy or downstream analysis. It works by guiding you through a structured profiling methodology and helping you interpret what the profile reveals about the data's fitness for purpose.

Profiling dimensions covered include column-level completeness (missing value rates, null distributions), uniqueness and distinctness analysis, value distribution summaries (min, max, mean, median, percentiles, skewness, kurtosis for numerical data; top-N value frequencies for categorical data), cardinality assessment, format and pattern consistency within text fields, cross-column correlation and dependency detection, temporal coverage and gap analysis for date fields, and row-level completeness across multi-column records. The assistant also helps you interpret profile results in the context of your data source and intended use.

Expected outputs include structured data profiling plans, Python code using pandas-profiling (ydata-profiling), pandas, and scipy for automated and custom profiling, interpretation guidance for profiling results, data quality issue prioritization frameworks, and data quality scorecards that summarize findings in a format suitable for stakeholder communication. The assistant also helps you design profiling as a repeatable, automated step in your data ingestion pipeline rather than a one-time manual exercise.

This assistant is the right starting point for any data cleaning project, any data migration quality assessment, and any vendor data onboarding process where you need to understand what you are receiving before you commit to using it.