MTPE Policy & Usage Guidelines – Trans Infopreneur Inc.

1. What MTPE Means at Trans Infopreneur

• Machine-generated translation is reviewed and corrected by a human editor
• The editor ensures:
  • Correct meaning
  • Sentences that sound natural and easy to read
  • Proper grammar
  • Use of approved terminology
• MT output is never delivered “as-is” without human review
• MTPE is used only with prior client approval

2. How Our MTPE Workflow Works

Step 1 – MT Output Preparation

• Files are prepared inside SDL Trados or Alchemy Catalyst
• Approved terminology and Translation Memories (TMs) are connected

Step 2 – Human Post-Editing

Human editors correct:

• Meaning errors
• Terminology errors
• Awkward or machine-style sentences
• Grammar and punctuation

Step 3 – Final QA Check

• Language completeness
• Terminology consistency
• Formatting and file integrity (files open and function correctly)

This ensures the final output is usable professional content — not raw machine text.

3. Where MTPE Is Usually Suitable

• Large-volume internal documentation
• Knowledge bases and FAQs
• Training material (non-regulatory)
• Software help content (non-safety-critical)
• Internal business communication

In such cases, MTPE offers:

• Faster turnaround
• Lower cost than full TEP
• Acceptable quality for internal or reference use

4. Where We Do NOT Recommend MTPE

• Legal contracts and compliance documents
• Court filings and regulatory submissions
• Medical device IFUs and patient-facing documents
• Clinical or safety-critical material
• Patent specifications and claims

For these, we strongly recommend full human TEP, as:

• Even small meaning mistakes are not acceptable
• Legal and safety consequences can be severe
• Approved technical and legal terms must be used from the beginning

5. Difference Between MTPE and Human TEP

6. Client Approval & Responsibility

• MTPE is used only after explicit client confirmation
• We clearly inform the client:
  • That the base text is machine-generated
  • What level of human correction will be done
• If a client later upgrades an MTPE project to full TEP:
  • We re-process the content under the full human workflow

7. Pricing Logic for MTPE vs Human TEP

• Human TEP is priced higher due to:
  • Full human translation
  • Editing and proofreading
  • Higher responsibility and risk
• MTPE is priced lower because:
  • Translation base comes from MT
  • Human effort is focused on correction and refinement

Both are still:

• Project-based
• Subject-area dependent
• Deadline dependent

8. What This Means for Agencies

• Transparent quality expectations
• Clear limits on where MTPE can and cannot be used
• Pricing clarity
• Protection against quality disputes
• Confidence in when to use MTPE and when to insist on full TEP

9. Machine Translation (MT) Engines We Use

• DeepL – strong for European languages
• Google Neural Machine Translation – broad language coverage
• Microsoft Translator (Azure) – enterprise IT and software MT

The choice of engine depends on:

• Language pair performance
• Subject complexity
• Client instructions

Regardless of the engine used, all MT output always goes through full human post-editing under our MTPE workflow.