How does iDBQuery understand a cryptic or obscure database schema?
iDBQuery understands cryptic schemas by enriching the raw structure with a semantic layer: it infers what abbreviated table and column names mean, pulls in definitions from data dictionaries and documentation, learns from the data itself, and lets you correct it — so obscure field names stop being a barrier to good answers.
Plenty of real databases were designed years ago by people who have since left, with names like `f_amt`, `dt_clsd` or `flg_3`. iDBQuery is built to work with exactly those, not just clean, well-documented schemas.
It builds understanding from several signals, layered by trust:
- Inference from names, types and the values it samples — recognising that `dt_clsd` is likely a close date.
- Documentation and data dictionaries you provide, which override guesses with authoritative definitions.
- Your own annotations and a business glossary, which sit at the top and win over everything else.
- Learned patterns from how queries actually resolve over time.
These are merged with a clear order of precedence, so a definition you confirm always beats an inferred one. iDBQuery can auto-enrich a newly connected source in the background, then surface what it's unsure about for you to confirm. The result is that even an opaque legacy schema becomes answerable in plain language — you don't have to rename columns, write documentation first, or remember which obscure field holds the number you need.
Updated 2026-06-22