<aside> 👩🏻‍💻

Category: AI Operations / Knowledge Retrieval / Agent Design

Tools Used: Claude API, Obsidian, Structured Documentation, Knowledge Architecture Frameworks

Timeline: 2026 – Present

Environment: Large-scale knowledge architecture containing 750+ interconnected records and 200+ published pages

</aside>

The Problem

As the archive expanded, information retrieval became increasingly difficult.

The challenge was no longer finding information.

The challenge was finding the right information, quickly, while maintaining context and preserving relationships between concepts.

Traditional search methods began producing diminishing returns as:

• note count increased
• metadata complexity expanded
• cross-references multiplied
• analytical frameworks evolved
• historical context accumulated across multiple years

The archive contained the answers.

Retrieving them efficiently became its own operational problem.

The Approach

Rather than treating AI as a general-purpose assistant, I approached the problem through role design.

The objective was not to create another chatbot.

The objective was to create a specialized archivist capable of navigating a complex information environment while operating within clearly defined boundaries.

The resulting system became Igor.