Week 2 Capstone · AI Engineering Roadmap

AI Model Usage Analyzer

Completed · Learning Project · 2026-08-23

A Python data pipeline that cleans, validates, and analyzes messy AI model API usage logs, flags IQR outliers, and exports analysis-ready results.

Tech stack: Python, Pandas, NumPy, CSV, JSON, Git

Source code (GitHub)

A Python data pipeline built as my Week 2 capstone. It transforms messy AI model API usage logs into analysis-ready data by cleaning and validating records, flagging unusual token usage, calculating overall and per-model metrics, and exporting clean CSV data and JSON summaries.

What I built

CSV ingestion and dataset inspection for shape, data types, missing values, duplicates, and descriptive statistics
A cleaning pipeline that normalizes model names, imputes missing token counts with the median, removes rows with missing latency and duplicate records, and enforces positive token and latency values
IQR-based token outlier detection that flags unusual observations for review instead of automatically removing them
Overall and per-model metrics for requests, token consumption, latency, and success rate, with cleaned CSV and JSON summary exports

Engineering lessons

This project reinforced that data quality has to come before analytics. Domain validation answers whether a measurement is valid, while statistical anomaly detection identifies observations that are merely unusual; keeping those decisions separate prevents legitimate workloads from being discarded. Separating processing, reporting, and export responsibilities also made the pipeline easier to understand and maintain.