Notaries & notarial staff · Workflow 03
Research a share-transfer question without pasting the deed file.
Create a hypothetical research question about representation, consent, and documents to check, without sending the dossier or an identification record to a general AI tool.
The moment this is useful
The task on your desk
A German GmbH share-transfer dossier raises a representation question: a seller intends to sign through an attorney-in-fact, and the articles contain a consent requirement. Before drafting, a member of the notarial team wants a structured list of points to research. The working file also contains identity documents, home addresses, and deed references.
Keep the facts the question depends on.
Keep for the research
- The jurisdiction and legal form: a German GmbH share transfer.
- The representation arrangement and the existence of a consent requirement.
- The distinction between known facts and issues still to be checked.
Remove or replace
- Names, dates of birth, home addresses, passport details, and signatures.
- Company names, register numbers, deed references, and identifying file paths.
- Purchase prices and ownership history unless essential to the specific question.
A worked prompt example
Fictional details · Edited by hand after reviewBefore review
For the transfer of shares in Linden Technik GmbH, HRB 123456, seller Jonas Berg will be represented by Lena Hart under a power of attorney. The articles require company consent. Deed draft UVZ 88/2026 includes the parties’ identity records. What should we check before drafting?
Reviewed prompt
Hypothetical German GmbH share transfer: the seller intends to act through an attorney-in-fact. The articles contain a company-consent requirement. List questions to research about the power of attorney, consent, and supporting documents. Separate statutory starting points from facts that require examination of the actual documents. Cite primary sources for independent verification; do not draft an execution-ready deed.
Use this workflow
- Write the legal issue as a hypothetical in your own words, outside the AI conversation.
- Use local detection to review any personal information still present.
- Manually remove register numbers, deed references, and transaction-specific details.
- Verify the suggested research points against statutes, professional sources, and the actual documents within your approved workflow.
What you should have at the end
A focused research agenda for the notarial team, rather than an AI-generated deed or an assessment of the actual power of attorney.
Local redaction is a preparation step. It does not establish compliance with notarial confidentiality duties or decide whether a particular AI service is appropriate.
Questions for this workflow
Can the model reliably detect every German identifier?
No. Current detection focuses on English. Review German names, legal abbreviations, register numbers, and deed references manually. A short hypothetical is preferable to pasting a complete dossier.
Why not keep the real company name if it is on a public register?
The company name can connect the question to a confidential transaction. Public availability of one fact does not make the surrounding dossier suitable to share.
Workflow context & further reading
These sources inform the workflow. The example and editing choices are Procurato’s illustration; they are not source-endorsed advice.
- GmbHG § 15 — transfer of shares — Statutory context for share-transfer formalities and additional consent requirements.
- Notare Ott/Zander: share-transfer preparation questionnaire — A practice’s example of the personal and transaction information collected for this work.
From example to your own work
Keep the context. Review the details.
Paste into the local redactor, review each detection, choose what to keep, and check the result before sharing it with your approved AI tool. Names are only part of the review: inspect confidential wording and identifying context too.
Free, no account. Initial model download: about 1.1 GB. Requires WebGPU and several GB of free memory. Detection focuses on English and can miss details. The examples here include manual editing; the tool does not automatically rewrite or summarise your text.