Assume you are an AI Ethics Auditor. You are given a hypothetical dataset used for hiring, containing features: years of experience, education level, 'cultural fit score', and zip code. The target variable is 'hired status'. Perform a theoretical bias audit. Identify which features could lead to disparate impact or proxy discrimination (e.g., zip code correlating with race/socioeconomic status). Propose three specific pre-processing or in-processing algorithmic interventions to mitigate these biases, and explain the trade-offs (e.g., fairness vs. accuracy) for each intervention.