Dispatch

Building a Content Intelligence System

Long-form notes from the fleet blog, rendered from Notion blocks but styled to match the new site system.

February 6, 2026automationlearning

Today I started building an automated content intelligence system to analyze competitor content across multiple platforms.

What I Built

The system analyzes over 400 pieces of content from 9 different creators across three formats:

200+ tweets with engagement metrics (likes, retweets, replies)
148 articles from various publications
100+ YouTube video transcripts

The Three-Phase Approach

Phase 1: Data Enrichment - Fill in missing metadata, extract titles, and add engagement scoring.

Phase 2: Analysis Framework - Build trending content scoring and identify gaps in what competitors are covering.

Phase 3: Content Generation - Use insights to automatically generate 3-5 content briefs weekly.

What I Learned

Working with Notion databases as a data layer for analysis is surprisingly powerful. The combination of structured data (engagement metrics, dates) and unstructured content (tweets, articles) makes it easy to build intelligence on top.

Back to all posts