FERC Form 1 Audit Explorer

Background

What recurring findings of noncompliance surface across FERC's audits of electric utilities, and could that history one day predict the issues in a new filing?

This explorer reads 71 FERC audit reports as one body of evidence instead of 71 separate PDFs. It shows which findings of noncompliance keep recurring, what FERC recommends about them, and how that differs between financial (FA) and non-financial (PA) audits. FERC publishes every final audit of electric utilities, ISOs, and RTOs (FY2015–present) at ferc.gov/audits — but one PDF at a time, so the patterns across the corpus stay invisible. The longer-term goal is an "audit-my-document" tool that flags likely issues in a new filing from this history.

How It Works

The tool surfaces what the reports already say and adds nothing on top. I vibe-coded the extraction pipeline, then pinned it down with a pytest suite so the parsing stays honest:

• Data source: FERC's public final audit reports (FY2015–present) from ferc.gov/audits — a scraped index of 71 reports, all rate-limited and cached on download.
• Pipeline: fetch → extract → structure → patterns → build. Each PDF is classified by FERC form and industry to isolate the Form 1 (electric) subset, then turned into per-page text.
• What it structures: the findings of noncompliance, the corrective recommendations, and the themes that recur across reports. Re-runs are idempotent — cached downloads are skipped.
• Traceable: every record carries the source URL and capture date of the official report, so any finding can be checked against the audit it came from.
• Scope: v1 structures the most recent electric reports; gas (Form 2) and oil (Form 6) audits are classified but not yet parsed.
• UX: a static vanilla HTML/CSS/JS site reading baked JSON, deployed to GitHub Pages.

[View the Live Explorer] | [View the Code]