Metadata manager: which parts can AI do?
A metadata manager needs Metadata management: 6 activities and 90 knobits, the smallest units of knowledge and skill worth learning and testing on their own.
- Mastered by the AI model
- Partly
- Where people lead
- Not tested
- Builds on (a prerequisite)
- Both mastered
- The activity’s knobits
Hover or tap a star to see the activity; what it builds on lights up.
How much of this job can AI do?
8knobits most AI models master
9that some models master
73where people still lead
0not tested yet
Of the 90 knobits the benchmarks cover, 9% are mastered by most of the 4 AI models tested. Mastering the knobits in a test is not the same as doing the job: someone still decides what to build, talks to the business, and answers for the result. But it shows where the work is shifting.
Activities, from most to least covered by AI
| Activity | Best AI model | Typical AI model (median) |
|---|---|---|
| Trace column-level lineage | 89% | 85% |
| Analyse the impact of a change | 90% | 84% |
| Describe the catalogue entries | 90% | 79% |
| Design the metamodel | 90% | 70% |
| Define business terms in the glossary | 92% | 67% |
| State the metadata requirements and strategy | 72% | 57% |
For your career: where people still lead
The knobits the fewest AI models showed they master. Being good at these is what a person adds next to AI tools.
- M-205 Recognises business meaning stored on a technical metadata class
- M-305 Recognises a definition too narrow for its concept
- M-107 Recognises inferred metadata published as approved
- M-102 Recognises metadata collected for no use case
- M-306 Recognises a general concept tied to one subject field
- M-407 Recognises two catalogue entries for one object
- M-309 Recognises a hidden definition of a second concept
- M-906 Explains the steps of building a metadata strategy
- M-902 Explains the goals and principles of metadata management
- M-905 Explains the types of metadata architecture
- M-204 Recognises a glossary term class unable to hold the term unchanged
- M-206 Recognises a metamodel class serving no use case
Where AI already helps the most
Knobits that almost every model tested masters: good candidates to hand to an AI assistant and review.
- M-352 Explains the substitution principle for definitions
- M-356 Explains essential and delimiting characteristics
- M-152 Explains business, technical and operational metadata
- M-313 Writes an intensional definition of a business term
- M-354 Explains extensional definitions and their completeness
- M-653 Explains change enablement for data changes
- M-251 Explains metamodel layers, classes and relationships
- M-308 Recognises an incomplete extensional definition
Choosing an AI model for metadata manager work
Best result for metadata manager work: gemini-4-argon (95% test score).
Per-activity results are on each activity page; all models are compared on the benchmarks page.
Test yourself
Skilirium's adaptive test measures you on the same knobits as the AI models. How a test works.