Extraction ✦ 85outlineCalibrate on your documents

Candidate span extract + code normalize

Who: Back-office / document ops. Steps: Regex/roster finds candidates → Choice picks span (+ classify currency/country) → code normalizes verbatim (email lower, E.164, Decimal). Expected effect: no invented digits outside the candidate set. Dedup: catalog card; Hands-on extract-and-validate keeps the recipe.

1

Treat this as an outline — adapt state and questions to your data.

2

Implement in code — call System One / Jev; compose answers yourself.

3

Gate on confidence — act, confirm, or escalate before side effects.

Outline sketch

Prompt
Audience: back-office / document ops.
Steps:
1) Regex/roster builds candidate spans.
2) Choice picks span (+ optional currency/country class).
3) Code normalizes verbatim (lowercase email, E.164, Decimal) — never invent digits outside candidates.
Expected effect: lower hallucinated values.
Cross-link: Hands-on [extract and validate](/guides/extract-and-validate/).

Needs access to: typesafe-sdkparsers

Who it's for

Back-office / document ops

Steps / how it's set up

Sources: pre_parsed_value_extraction cookbook · Hands-on extract-and-validate.

Pipeline: find candidates → pick → normalize in code. Cross-link Hands-on extract-and-validate.

Sources (wave-2 deepen)

Expected effect

Jev selects among known candidates; deterministic code owns the final string.

Unofficial outline for learning. Paraphrased from public docs and tutorials — not a production recipe. Review sources before you automate anything.