# The conversion that _finishes._

Hand a twenty-year estate of FrameMaker, Word, RoboHelp, InDesign, and scanned PDF to a pipeline that actually finishes — clean DITA, the duplicates already collapsed, every topic traceable to the source it came from. The migration that used to stall in a typing pool ships on a schedule. Because the duplicates collapse during the conversion — not months after go-live — the new repository starts smaller and cheaper to translate, search, and feed an AI pipeline. What lands is a clean DITA project your team can build on, not a re-keyed copy of the estate you started with.

## Source Formats We Convert From

One engineered pipeline, every legacy estate — not a different tool and a different team for each format.

- Microsoft Word
- HTML
- Custom XML / DTD
- Excel
- FrameMaker
- SGML
- PowerPoint / PPTX
- RoboHelp
- Confluence
- AsciiDoc
- Legacy DITA (1.0–1.2)
- DocBook
- MadCap Flare
- Markdown
- QuarkXPress
- InDesign
- Google Docs
- AuthorIT
- PDF (OCR)

## How Conversion Works

Our conversion methodology combines proven migration tools with deep DITA expertise to handle the heavy lifting. You focus on quality.

1. 01

### Source Document

- Inconsistent heading styles
   - Manual numbered lists
   - Embedded images at random DPI
   - No metadata or taxonomy
   - Duplicate content across files
2. 02

### DITA Output

- Clean topic types (concept, task, reference)
   - Semantic <ol> and <steps> markup
   - Normalized images, consistent DPI
   - Rich metadata and SubjectScheme keys
   - Deduplicated with conref / conkeyref reuse
3. 03

### AI-Ready Content

- Chunked topics optimized for RAG retrieval
   - Schema.org and JSON-LD metadata mapping
   - Semantic labels for LLM grounding and citation
   - Embedding-friendly short descriptions per topic
   - Knowledge-graph nodes from the subject-scheme taxonomy

## Automatic Topic Deduplication

During conversion, we analyze every paragraph and span across your entire document collection using industry-proven migration tools and structured content analysis. Exact-match duplicates are identified and consolidated into reusable warehouse topics with conref pointers — saving you from discovering redundancy months later.

Typical result: **15–30% of topics** identified as duplicates and consolidated in the first pass.

## The Modernization Journey

We don’t just move files; we re-engineer information architecture. From unstructured chaos to intelligent, reusable XML.

1. 01

### Analyze

Dedup and audit the legacy source for recoverability
2. 02

### Model

Define the target information architecture and reuse model
3. 03

### Transform

Automated, validated XML conversion, format by format
4. 04

### Enrich

Add semantic metadata, taxonomy, and reuse keys
5. 05

### Go Live

Validate, baseline, and hand off with full provenance

- ### Legacy Conversion

We build custom parsers for FrameMaker, InDesign, Word, and RoboHelp to extract maximum structure.

- ### Metadata Strategy

Applying taxonomy tags during migration for future faceted search and dynamic filtering.

- ### Quality Assurance

Automated schematron validation ensures every migrated topic adheres to your new content model.

## Sample Content Assessment

Send us a 20-page sample — FrameMaker, Word, RoboHelp, InDesign, scanned PDF, anything. We’ll return a conversion feasibility, content recovery rate, and migration-effort estimate within two business days.