Python Data Analysis on a Provided Dataset (Capacity Management Context)

You are given a dataset (file or table). Using Python, perform an analysis and produce reproducible outputs (code + conclusions). The dataset relates to cloud capacity management / optimization. Tasks (implement as requested by the interviewer; cover at least the first three) Load and clean the data: handle missing values, outliers, and duplicates. Compute key metrics: e.g., utilization, peak/mean, P95/P99, over/under-provisioning ratios, cost-related metrics. Identify bottlenecks and drivers: use aggregations, group-bys, correlation/regression, or other interpretable methods to explain low/high utilization. Provide optimization recommendations: propose ways to reduce waste or cost under SLA/reliability constraints and explain trade-offs. (Optional) Build a simple forecast/planning step: estimate future demand and recommend a capacity buffer. I/O and constraints Input: a dataset provided by the interviewer. Output: runnable Python code (plus a short written summary). Constraints: clearly state assumptions, edge cases, and metric definitions.