AI Native Engineering
AI-Native Engineering Assessment
8 questions · under 3 minutes · 32 points maximum
0 of 8 answered
01
Where does AI sit in your delivery lifecycle?
Code completion and generation only
1
Codegen plus testing or documentation
2
Requirements through code through test, with gaps
3
Specification through verification as a continuous pipeline
4
02
Where is human judgment mandatory in your AI-assisted work?
Not defined; reviewer discretion
1
Informal norms, unwritten
2
Written policy for high-risk changes only
3
Documented mandatory-review points, enforced in the pipeline
4
03
How structured are your requirements before AI touches them?
Prose tickets and conversation
1
Templated tickets, human-oriented
2
Structured acceptance criteria on most work
3
Machine-actionable specs with explicit contracts and constraints
4
04
How coupled is your pipeline to a single AI vendor?
Fully dependent on one vendor's tooling
1
One primary vendor, others experimentally
2
Multiple tools, integration is manual
3
Vendor-abstracted pipeline; tools are interchangeable components
4
05
In a two-week sprint, how many merged PRs are produced by agents under a single developer's supervision?
Fewer than 10; essentially human-authored pace
1
10–40; agents contributing at scale, developer still authoring
2
40–120; developer operating primarily as supervisor
3
120+; developer running a fleet, throughput bounded by verification
4
06
If an AI-originated defect shipped, could you trace its provenance?
No
1
Reconstructable manually with effort
2
Partial traceability on some workstreams
3
Full traceability: prompt, model, review, approval — auditable
4
07
How often do you replace the underlying LLM, and does output quality hold?
Never changed it
1
Changed once; quality impact not measured
2
Change periodically; quality checked informally
3
Change regularly against a measured baseline; quality holds
4
08
What happens to what you learn on a project?
Stays with individuals
1
Documented but rarely reused
2
Shared assets exist, adoption is uneven
3
Codified into reusable assets that measurably improve the next engagement
4
See my score
AI Native EngineeringAI Native EngineeringAI Native Engineering