AI Generates 214-Page Cited Wiki from 45k Enron Emails

·2026.10.23 21:02

Key point

A prototype system processed 45,047 Enron emails to create a 214-page entity wiki where 93% of sampled facts were verified as fully supported by source documents.

Details

An AI pipeline processed 45,047 Enron emails to generate 214 pages covering companies, regulators, and banks, with 49,401 facts each citing a specific source email.
A held-out audit by a second AI found that 93% of sampled facts were fully supported by their emails, with 0 of 500 sampled facts wrong at their core.
The system identifies temporal validity, expiring facts on the date the email indicates they stopped being true, and allows users to browse knowledge by specific dates.
The project uses the public Enron corpus to demonstrate a 'company brain' that tracks changing relationships and facts, distinct from static CRM fields.

generative-ai knowledge-graph entity-extraction enron fact-verification

Read originalShare on Threads

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.

Rust off Rails: An opinionated web stack for the AI era using Markdown conventions instead of code