Each person appears once.
Every pair-task variable competes for the same roster. A locally perfect choice can consume the only evidence another challenge needs.
Pair every strength to the right challenge.
Upload real evidence, solve all pairs together, and inspect every covered or missing requirement. Age is never converted into ability.
name, age, skills
Skills use semicolons inside a quoted field. They must be observed or self-reported; age never creates one.
challenge, required_skills
Each challenge receives exactly one pair. Required skills also use a quoted semicolon list.
Final tie-break
Complete tasks—
Requirements—
Age spread—
Model size—
Complete challenges—
Named coverage—
Age spread tie-break—
Optimized seeding
Waiting for validated CSV
One global solve, not three greedy picks.Load the sample or paste an even roster. The solver keeps every person and task in one exact model.
The objective order is strict: complete tasks first, named coverage second, optional age spread last.
Why the global assignment changes the answer.
Finish as many challenges as possible.
A task is complete only when the pair's explicit skill union contains every named requirement. No age coefficient or hidden aptitude score exists.
Preserve the win before breaking ties.
The solver locks the complete-task optimum, then maximizes covered requirements. Age spread is optional and acts only after both higher-priority outcomes are fixed.