Your role
AI skills for data analysts
Analysts now use AI to draft SQL, explain unfamiliar code, clean messy files, and write the paragraph that sits above a chart. The output looks finished. A query runs and returns rows. A summary reads like an analyst wrote it. That makes errors harder to see, because a wrong number looks the same as a right one.
This guide breaks the work into the six skills Emeri measures. Each skill has its own page with tasks from analytics work, a description of the skill at each band from E1 to E4, the mistakes analysts make most often, and a worked example.
- 1
See where you stand. A free check of the six skills, about fifteen minutes.
- 2
Build the skills. Free guides and courses that fit data work.
- 3
Prove it. One examined sitting, then a grade anyone can verify.
Where AI shows up in the work
- Writing and reviewing SQL
- Drafting queries, translating logic between dialects, and explaining a colleague’s query before you change it. A query that runs is not the same as a query that answers the question.
- Cleaning and preparing data
- Parsing inconsistent dates, standardizing category labels, and writing transformation code in a notebook. AI writes the code fast. You still decide what counts as a valid value.
- Explaining results
- Turning a metric change or a dashboard into a short narrative for a stakeholder. The narrative must say only what the data supports, including what it cannot tell you.
- Documentation
- Describing tables, pipelines, and metric definitions so the next person can trust and rerun the work. AI drafts well from code. It cannot know the business rules nobody wrote down.
Step 1
See where you stand
The free check grades all six skills in about fifteen minutes, and shows which to work on first. AI use is encouraged. No account is needed, and nobody sees the grade but you.
Want to know more first? Read about the check.
Step 2
Build the skills
Start with the skill your check showed as weakest. Every guide, course and template here is free.
The six skills, for data analysts
Each guide shows what the skill looks like for data analysts at every band from E1 to E4.
Asking well
Turning a business question into a request with a metric definition, a grain, a time window, and an output format.
Read the guide →Giving good context
Supplying the schema, sample rows, metric definitions, and known data quirks the model cannot see.
Read the guide →Checking the work
Confirming that every query, figure, and statistical claim traces back to the source data.
Read the guide →Putting it together
Turning one-off prompts into repeatable steps for queries, notebooks, recurring reports, and documentation.
Read the guide →Getting real work done
Taking work from a draft query to a finished analysis a stakeholder can read, trust, and act on.
Read the guide →Keeping it safe
Protecting personal and confidential data, following governance rules, and refusing to present weak evidence as fact.
Read the guide →Courses that fit data work
Each course is written for every job. Here is why each one fits yours.
- Advanced · 6 lessons
Building and automating
Its pipelines lesson covers strict output formats, logging every input and output, and sampling results on a schedule.
Start the course → - Free · 7 lessons
Presentations people care about
Its evidence lesson shows which charts readers judge most accurately, and puts a source line under each one.
Start the course →
Templates for data work
Fill-in forms you can copy, each with a filled example.
AI context brief template
A one-page project background you write once and paste at the start of each conversation.
Open the template →AI output verification checklist template
A form for deciding how one recurring deliverable gets checked before it ships.
Open the template →AI workflow map template
A form for laying out a multi-step AI task, with review placed where errors would spread.
Open the template →Step 3
Prove it
When you are ready, the certification examines the same six skills in one sitting. AI is allowed, as it is in the check. Pass, and your grade gets a public page that anyone can use to verify it. One retake is included if your first sitting falls below the bar.
Get certified — $129Questions
What AI skills do data analysts need?
Emeri measures six: asking well, giving good context, checking the work, putting it together into workflows, getting real work done, and keeping it safe. In analytics, checking the work and giving good context tend to separate reliable analysis from plausible analysis. AI output in this field fails most often on schema assumptions and unverified numbers.
Will AI replace data analysts?
Emeri does not make predictions about jobs. AI changes which parts of analytics work take time. Writing a first draft of a query or a summary is faster. Defining the question, checking the numbers, judging what the evidence supports, and protecting the data still sit with the analyst. That judgment is what these guides and the Emeri assessment focus on.
How can I tell what level my AI skills are at?
Compare the band descriptions on each skill page with how you work today. For a measured result, take the free Emeri check. It takes about fifteen minutes and gives an indicative band across all six skills.
Are these example tasks from the Emeri exam?
No. Emeri wrote these tasks for this guide, and none of them appear in the exam. The certification has its own scenarios, which examiners grade against the same six skills.
Hiring for this role? The AI interview questions for data and analytics candidates explain what a strong answer and a weak answer show for each skill.
Start with step 1
The free check takes about fifteen minutes. Your result shows your strongest and weakest skills, so you know which guides to read first.