Jobs › Computer and Mathematical

AI exposure report

Will AI replace Data analyst?

Many of your tasks are within reach of AI. The role will change, so it pays to act now.

Task exposure to today's AI

61/100

High exposure

This is the occupation's typical task mix. Your own score can be up to 30 points higher or lower depending on how you work: routine share, people contact, judgment and how much you already use AI.

Get your personal score

Will AI replace Data analyst?

Not entirely. Today's AI tools can take on about 61% of the typical work in this occupation, mostly: cleaning and preparing data The parts that stay human include: Asking the right question and understanding the business context; Knowing data quality and its pitfalls; Telling stories with data and influencing decisions.

People who learn to use AI for the automatable parts keep the edge. The 3-phase plan below shows where to start.

Which tasks AI can take on

The largest tasks in this occupation, their share of the working week, and how much of each today's AI tools can do.

TaskAutomatable
Cleaning and preparing data 80%
SQL queries 80%
Reports and dashboards 75%
Interpreting results and making recommendations 35%
Working with business users 20%

Your human edge

  • Asking the right question and understanding the business context
  • Knowing data quality and its pitfalls
  • Telling stories with data and influencing decisions

AI tools to try

ChatGPT/Claude data analysisCopilot in Power BIText-to-SQL

Related roles

  • Analytics engineer
  • AI/ML analyst
  • Product analyst

A 3-phase AI learning plan

Phase 1

AI as an analysis accelerator

  • Let AI write SQL and Python – and learn to check it.
  • Exploring data with AI (uploading files, charts).
  • Automatic summaries of reports.

Project

Take a public dataset (e.g. from a national statistics office) and produce an analysis + written report with AI in 2 hours.

Phase 2

Semantic layer and text-to-SQL

  • dbt and a semantic model so that AI answers are correct.
  • Text-to-SQL solutions and their limits.
  • Python + an LLM API for classifying data.

Project

Build an internal 'ask the data' tool that answers 10 typical questions.

Phase 3

ML and predictive analytics

  • Basic ML models (regression, classification) with scikit-learn.
  • Metrics and experiment design (A/B tests).
  • Model explainability.

Project

A customer churn prediction model + an explanation for a business user.
Take the free 11-question test