Methodology
This page explains how Se3r.today collects prices, assigns freshness states, and uses fallback or synthetic paths when upstream sources fail.
Collection model
The platform collects global and local gold signals, core FX references, and public bank-rate rows from multiple public sources depending on the market surface.
Each data family has its own refresh cycle, with recent snapshots stored in PostgreSQL and Redis so public pages can stay fast and resilient.
- Crypto prices: fast cycle around every 30 seconds.
- Gold prices: primary cycle around every 60 seconds.
- Bank-rate and FX reference data: cycle around every 30 seconds.
Freshness and trust signals
Every important market response exposes an explicit state such as live, degraded, stale, fallback, or unavailable, alongside the most recent observed timestamp.
The state is derived from snapshot age, upstream success or failure, and fallback markers rather than a single front-end guess.
Fallback behavior
When a primary source fails, the system attempts a secondary source or the latest trustworthy stored snapshot. In extreme bank-board failures, clearly marked synthetic rows may be published to avoid showing a blank board.
Synthetic or estimated rows are explicitly labeled and should not be interpreted as direct live bank quotes.