About signal.library Below is a working prototype of signal.library, a personal intelligence system that turns my ongoing research into a searchable, compounding knowledge base. While most uses of AI in landscape research focus on the broad accumulation and synthesis of outside sources, I was inspired by the idea of leveraging AI to help me sift…

About signal.library

Below is a working prototype of signal.library, a personal intelligence system that turns my ongoing research into a searchable, compounding knowledge base. While most uses of AI in landscape research focus on the broad accumulation and synthesis of outside sources, I was inspired by the idea of leveraging AI to help me sift through and enrich a selective, personally curated information archive.

The current prototype was built using a small corpus of signals focused on the future of influence and attention online, which is a domain I’ve been tracking closely throughout the newsletters, panels, articles, and podcasts that make up my regular information diet. Every signal in the library is something I personally encountered, saved, and found compelling enough to write up in my archive. The interpretations are entirely my own.

My hope is to expand this tool over time by adding more signals to the existing topic area corpus, and eventually building out parallel libraries for other topic areas I’m actively researching.

A bit about my process: To build this prototype, I organized existing notes and signals from my personal archive into structured markdown files in Obsidian. I had used Obsidian in the past to build a repository of notes during grad school, and it was a natural fit here. I then used Claude Code to build a weekly signal processing workflow: it asks me pointed questions to push my thinking, records my verbatim responses alongside the original signals, and surfaces connections and further provocations across the library. Finally, Claude Code helped me develop a tailored interface that acts as a dynamic, working repository.