Document date
Date displayed on the letter.
Generate an editable preliminary medical device manufacturer-to-local-representative appointment letter from pasted entity, product, market, responsibility, and signatory facts.
Formatting does not matter. Pi maps reliable facts into this document’s fixed JSON; deterministic code writes the branded editable Word file.
Each field has its own description in the agent prompt and output schema. A key is always present in the JSON, but its value is null when the pasted material does not support it.
Date displayed on the letter.
Exact appointing manufacturer.
Exact appointed entity and supplied identifiers.
Market, authority, and exact role.
Exact products, models, accessories, or registrations.
Only the responsibilities supplied by the user.
Start, duration, notice, and regulatory-transition terms when supplied.
No signature is applied.
Acceptance signatory when supplied; no signature is applied.
Structure repeated company, product, model, approval, certificate, and evidence facts first.
Use current stored ANVISA and COFEPRIS source artifacts with form-specific JSON.
Turn an evidence index into a source-backed missing-attachment worklist.
Check repeated facts, files, signatures, portal mapping, and release blockers.
No. Pi must first return the complete document-specific JSON contract. Deterministic code validates that object and writes the editable DOCX.
The field stays blank and appears in the review list. The workflow is instructed not to guess names, identifiers, dates, powers, classifications, standards, certificates, or authority language.
No. The output is an editable preliminary draft. Execution, acceptance, local wording, signatures, notarization, apostille, legalization, translation, and submission remain human responsibilities.
Yes. The result is a native DOCX draft intended for review and editing in Microsoft Word or another compatible editor.