A non-developer built a RAG-style AI tracking dashboard with Claude Code
A management consultant spent about six months building a live dashboard for tracking AI developments with heavy help from Claude Code. The system gathers information from sources such as papers, lab announcements, GitHub, Reddit, and policy feeds; it started with 16 sources and now uses 21.
It groups related events by embedding , turns multiple sources into combined stories, and shows them on a live dashboard at techvenue.com. A subscriber feature later added , where a person types a question and gets an answer based on the past 60 days of stored stories.
A technically minded friend identified the system as RAG before the builder knew that term. The main concern is whether the design is sound or just an improvised system that happens to work.
Key points
- Claude Code helped a non-developer build an AI trend dashboard.
- The system collects from 21 sources, including papers, lab news, GitHub, Reddit, and policy feeds.
- It groups similar events using embedding and turns multiple sources into combined stories.
- Subscribers can ask over 60 days of stored stories.
- The item is more useful for RAG than for direct reduction.