Your role
AI skills for software engineers
Engineers now ask AI to explain unfamiliar code, draft tests, suggest fixes, and write first versions of design docs. The difference between engineers shows in how they use it. Some accept a plausible diff and find the bug in production. Others give the model the right files, read every line it returns, and run the tests before they open a pull request.
This guide splits that difference into the six skills Emeri measures. Each skill has its own page. Each page lists tasks from engineering work, describes the skill at each band from E1 to E4, names common mistakes, and walks through 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 engineering work.
- 3
Prove it. One examined sitting, then a grade anyone can verify.
Where AI shows up in the work
- Writing and changing code
- New functions, refactors, migrations, and dependency upgrades. AI writes code quickly. The skill is getting code that fits the codebase’s conventions and does not break callers you did not mention.
- Reading and debugging
- Explaining an unfamiliar module, tracing a stack trace, or forming hypotheses about a flaky test. The risk is a confident explanation of code the model has not seen.
- Tests and review
- Drafting unit tests, finding edge cases, and giving a pull request a first pass. A test that passes is not proof the code is right. The engineer still decides what the test should assert.
- Engineering writing
- Design docs, incident write-ups, runbooks, and pull request descriptions. AI can structure the document. The decisions, the timeline, and the root cause have to come from the engineer and the evidence.
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 software engineers
Each guide shows what the skill looks like for software engineers at every band from E1 to E4.
Asking well
Turning a coding or debugging need into a request with a goal, constraints, and a definition of done.
Read the guide →Giving good context
Supplying the relevant code, conventions, versions, and constraints the model cannot see on its own.
Read the guide →Checking the work
Catching invented APIs, subtle logic errors, and tests that pass without proving anything.
Read the guide →Putting it together
Turning one-off prompts into repeatable steps for features, reviews, upgrades, and incident follow-up.
Read the guide →Getting real work done
Taking work from generated code to a change that is tested, reviewed, documented, and merged.
Read the guide →Keeping it safe
Protecting secrets, proprietary code, and user data, and checking the licensing of generated code.
Read the guide →Courses that fit engineering work
Each course is written for every job. Here is why each one fits yours.
Templates for engineering 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 workflow map template
A form for laying out a multi-step AI task, with review placed where errors would spread.
Open the template →Team AI-use policy outline
An outline a manager can fill in to set clear AI rules for one team.
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 software engineers 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 engineering, these guides give particular weight to checking the work and giving good context. A model that lacks the codebase’s context can write code that looks right and is not, and only careful review finds it.
Will AI replace software engineers?
Emeri does not make predictions about jobs. What we can say is that AI changes which parts of engineering work take time. Writing code is faster. Decisions about design, correctness, security, and what to build still rest with the engineer. That judgment is what these guides and the Emeri assessment focus on.
How can I tell what level my AI skills are at?
Read the band descriptions on each skill page and compare them 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. The tasks on these pages are written for learning and are never exam questions. The certification uses its own scenarios, graded against the same six skills.
Hiring for this role? The AI interview questions for engineering 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.