Japan Business Data Normalizer

by Kaito IshizukaUpdated Jul 18, 2026

Clean Japanese CRM company names and phone numbers into stable JSON before matching, review, or import.

data cleaning
crm
etl
+7
|

Clean inconsistent Japanese company names and phone numbers before CRM import, matching, ETL, sales-list processing, or agent workflows. One business-record call returns both normalized fields and preserves a usable company result even when the phone number needs review.

Verified hosted quick start

npx mcpize@latest login
npx mcpize@latest run japan-business-data-normalizer-mcp/normalize_japanese_business_record \
  '{"company_name":"㈱サンプル商事","phone_number":"03-1234-5678"}' --json

The server uses deterministic local rules and bundled Japan numbering metadata. It normalizes Unicode width and whitespace, recognizes supported legal-entity forms only at safe name boundaries, generates a non-unique search key, parses domestic and explicit +81 phone formats, keeps extensions separate from E.164, and preserves partial success.

Every response includes stable status, warning codes, change codes, schema version, algorithm version, and normalization confidence. Workflows can accept normalized results and route partial or unresolved results to review without parsing prose. The same JSON is available as structured content and text fallback.

This server does not use an LLM, external API, database, or persistent application storage. It does not verify company existence, registered legal names, corporate numbers, record identity, phone ownership or reachability, addresses, non-Japanese numbers, or short service codes.

日本語

表記揺れのある日本企業名と日本国内の公開電話番号を、CRM取込、名寄せ候補生成、営業リスト、ETL、AIエージェントで扱いやすいversion付きJSONへ決定論的に正規化します。複合Toolは電話番号を正規化できない場合も会社名の成功結果を保持します。公開版は上記のmcpize login後、mcpize runで実行できます。

すべての結果にstatus、confidence、warning code、change code、schema version、algorithm versionが含まれます。LLM、外部API、外部DB、applicationによる入力保存は使用しません。

法人の実在・登記上の正式名称・法人番号・同一企業性、電話の所有者・到達性、住所、海外番号、short codeは確認・処理しません。search_keyは候補生成用であり一意キーではありません。