External Intelligence (SEO & Branding)

The intelligence module represents the “Eyes and Ears” of the platform. It moves beyond internal datasets to crawl, scrape, and aggregate the “Digital Footprint” of a project and its competitors.

🔍 SEO Auditing: The BFS Graph Crawler

Our SEO engine uses a Breadth-First Search (BFS) graph traversal strategy. It doesn’t just look at a homepage; it maps the entire “Interconnectedness” of a domain.

The Reasoning Behind the Crawl

• Depth vs. Performance: We default to a depth of 20 levels. This is calibrated to capture all primary conversion pages and sub-hubs without overwhelming the target server or our own queue buffers.
• The Issue Hierarchy: The crawler classifies issues into Errors (Critical, like 404s), Warnings (Performance, like missing Alt tags), and Notices (Optimization, like long titles). This hierarchy allows the frontend to render the “SEO Health Score” as a prioritized roadmap rather than just a list of bugs.

👂 Brand Listening: Native Social Aggregation

The Media Intelligence sub-module performs sentiment-weighted brand tracking across the open social web.

Multi-Source Normalization

Social platforms (Twitter, Reddit, YouTube) use wildly different data schemas. Our Media Aggregator serves as a translation layer:

• Snscrape Integration: For X and Reddit, we attempt to use native scrapers which capture not just the “Post,” but the metadata (Likes, Retweets) that dictates the Influence Weight of a mention.
• The Weighted View: Not all mentions are equal. A post with 10k likes contributes more significantly to the “Sentiment Trendline” than a post with 0 engagement. This logic is baked into the aggregator before it even hits the Intelligence Engine.

🌓 Sentiment Reasoning

Every crawled mention is passed through a two-stage classifier:

1. Lexical Pass: Fast, polarity detection based on keywords.
2. Contextual Pass: For high-engagement posts, we use a lighter GenAI pass to detect Sarcasm or Slang that typically trips up standard sentiment models. This ensures that the branding dashboard accurately reflects the “Vibe” of the public conversation.