Clean research data consistently
Apply the same rule to every survey response instead of relying on invisible manual edits.
Python for UX researchers
Learn enough code to investigate, build and challenge AI.
You do not need to become a software engineer. You need enough technical fluency to work with evidence, understand system behaviour and test whether an AI experience deserves users’ trust.
Why Python belongs in UX research
Code makes methods repeatable and assumptions visible
Spreadsheets are useful, but Python becomes valuable when a study contains repeated cleaning rules, many sessions, custom measures or AI outputs that need consistent evaluation.
Calculate defensible measures
Reproduce task success, time, comparison and group-level results when the dataset changes.
Evaluate unpredictable systems
Test retrieval, confidence, failure routes and quality criteria across many AI interactions.
The learning sequence
One portfolio project, built in five stages
Each stage teaches isolated ideas first, confirms learning, then contributes those ideas to a small retrieval assistant you can download and explain.
Absolute beginner
Values, input, decisions and the project’s intended audience.
Building blocks
Lists, loops, functions, records and the working prototype.
UX evidence
Clean sessions, task measures and meaningful comparisons.
AI foundations
Tokens, intent, similarity, retrieval and evaluation.
Responsible bridge
Confidence, oversight, user groups, limitations and launch decisions.
Free practical tools
Start with a real AI UX decision
AI UX readiness check
Identify the smallest useful starting point across code, research and responsible interaction design.
Check your readiness →Chatbot UX rubric
Score one assistant task against eight observable experience criteria.
Use the rubric →Confidence fallback tester
Explore when an AI interface should answer, clarify or hand off.
Test the routes →Questions beginners ask
What this path does—and does not—promise
Do UX researchers really need Python?
Not for every role or study. It is most useful for mixed-methods work, repeated analysis, experimentation, AI evaluation and collaboration with technical teams.
Do I need to install anything?
No. Python runs in a private browser sandbox. Code and progress currently remain in your browser.
Will this make me an AI engineer?
No. It provides practical coding and evaluation foundations for human-centred AI work. Building production models requires substantially more computing and mathematics.
What will I have at the end?
A working retrieval assistant, its source file, a small knowledge base, evaluation logic and the foundations of a portfolio case study.
Write your first line. Build toward your own thing.
The examples are supplied; the final audience, knowledge, risks and evaluation decisions belong to you.