Quantamental

Quantamental answers a hard question quantitatively: how credible is a given central bank? It scores inflation-mandate adherence, monetary-policy stance, communication and geopolitical pressure across seven weighted dimensions, then surfaces the signals as a live macro product.

Under the hood it is a Dagster-orchestrated Python pipeline of 49 assets over 1.67 million point-in-time observations, LLM-based analysis of central-bank communication, a FastAPI service and a PostgreSQL warehouse. Every superseded value is kept (213,000 so far), so a score can be re-run for any past day. It is the production version of the macro-credibility work shown in my consulting case studies, and its engineering case study is public at quantamentry.com/case-study.