LOKAL — Preparing a sample for Document AI assessment
====================================================

1. DOCUMENTS
Choose a representative set of both typical and difficult documents.
Include the formats, languages, older typed documents, tables, numbers and
poor-quality scans present in your collection. Identify material that needs
human review.

2. OBJECTIVE
Describe what you need: OCR, corrections, export, questions about selected
documents, collection-wide search, transcription or field extraction.
Collection-wide search and transcription are separate implementation scopes.

3. VOLUME
Estimate your historical collection, incoming document volume and priorities.
State your expected processing time and number of users.

4. SYSTEMS AND INFRASTRUCTURE
Describe your existing DMS, CRM, ERP or archive and available integrations.
Include network requirements, offline operation, permissions and hardware.

5. SUCCESS CRITERIA
Identify the results that matter: text, numerical values, specific fields,
processing time and the review workflow. Quality is assessed on your sample.

6. CONTROLLED SHARING
Share material only after agreeing the scope and a controlled way to provide it.
Confidential files can remain on your infrastructure.

ASSESSMENT OUTCOME
An agreed pilot scope, quality criteria, hardware requirements,
an integration plan and clearly documented limitations.