Raw LLM Responses

Here's a crazy idea. The idea behind getting companies the best hire possible is, at the end of the day, to drive success for them, right? So instead providing the "success" rate of individual hires, the databset should be of the company's performance metrics and the resumes of all its employees. Then, the ML algo optimizes not for success rate of any given resume, but for the entirety of the company's personnel make-up. The reason I think this would be more effective in achieving a balance of diversity AND performance is because many studies show diversity is linked to corporate success, but you'd still allow for adjustment by industry / KPI. In other words, treat the problem holistically like diagnosing a disease, not treating a single symptom. If you have brain cancer, the best aspirin in the world is not going to make you healthier.