Offshore logic
Personnel movements, rotations, POB and accommodation must remain operationally credible and connected.
BENCHMARK 05 / SOFTWARE ENGINEERING / 2026
Offshore ATS.
One prompt. Eight systems.
Can AI translate a detailed offshore workforce-management specification into a functional browser application without frameworks, databases, APIs or iterative refinement?
THE TEST
Function first.
Inspect the system.
Eight leading AI models received the same specification and independently built an Offshore Attendance Tracking System. Each output had to run as a self-contained browser application with no external frameworks, databases or APIs.
The benchmark tests whether a model can turn interconnected operational requirements into coherent software—not whether it can produce the most polished interface.
PERSONNELROTATIONSPOBACCOMMODATIONLOGISTICSCOMPETENCYREPORTING
One system state
Changes should remain aligned across records, dashboards, logistics, competencies and reports.
Working software
The application must make complex workforce information understandable and usable—not merely display sample screens.
THE OUTPUTS
Test all eight.
One specification. Eight independent first-pass applications. No iterative refinement and no declared winner.
01
OpenAI
Launch application ↗ 02Perplexity
Launch application ↗ 03Gemini
Launch application ↗ 04DeepSeek
Launch application ↗ 05GLM
Launch application ↗ 06Qwen
Launch application ↗ 07Claude
Launch application ↗ 08Mistral
Launch application ↗IMPORTANT
Return to all benchmarks →
All applications are first-pass prototypes using fictional data. They are not production-ready systems and should not be used for live personnel, POB, logistics, accommodation or competency management.