Automating Legacy Content Conversion with AI

Background

For more than 40 years, iCEV has provided Career and Technical Education (CTE) curriculum used by educators nationwide. Their platform includes multimedia lessons, assessments, vocabulary exercises, worksheets, and certification preparation materials across disciplines such as agriculture, health science, business, and law enforcement.

Over decades of curriculum development, iCEV accumulated thousands of instructional documents in multiple formats, many created in Microsoft Publisher. As their platform and publishing standards evolved—and with Publisher approaching end-of-life—the organization needed a scalable way to migrate this legacy content into standardized Microsoft Word templates without compromising instructional accuracy. To solve this challenge, iCEV partnered with IntelliTect to design and build an automated conversion tool powered by AI.

Challenges

The primary challenge was scale combined with formatting complexity. iCEV’s content library contained thousands of documents spanning 23 distinct document types, including lesson materials, assessments, answer keys, and classroom activities. Each document type had unique layout requirements and formatting rules, making simple script-based conversion impractical.

The legacy format added further complications. Microsoft Publisher provides limited programmatic access and no reliable export options that preserve layout fidelity. Converting files directly to Word often distorted images, tables, and structured activities such as matching exercises or line-art worksheets.

Maintaining instructional integrity was critical. Many documents contained structured educational content like diagrams, numbered exercises, answer tables, and graphics that needed to survive the migration intact. Manual conversion was technically possible but would have required significant time from the content team and introduced consistency risks across thousands of files.

iCEV needed a solution capable of interpreting document structure, reconstructing content accurately, and processing large volumes of files efficiently.

Solutions

IntelliTect developed a C# command-line tool that combines Office automation with Azure-based AI services to automate the conversion pipeline. The system extracts content from legacy Publisher files, interprets the structure using AI, and reconstructs the content inside standardized Word templates.

Key components of the solution included:

1. AI-Driven Document Classification

The system automatically determines which of iCEV’s 23 document types each file represents by analyzing filename patterns, visual signals from the first page, and structural characteristics within the document. This allows the tool to route each file through the appropriate template and formatting rules.

2. Publisher Content Extraction and Transformation

Because Publisher lacks reliable export tools, the application retrieves document content using Publisher COM Interop. The content is exported as HTML—often messy but structurally recognizable—then processed with AI to reconstruct the document into a clean structure suitable for Word templates.

3. AI-Assisted Content Reconstruction

Using Azure OpenAI, the system restructures document content to match the appropriate template while preserving instructional elements such as images, tables, lists, and worksheet layouts. Large documents are processed in smaller sections to maintain language model performance and accuracy.

4. Batch Processing for Large Document Sets

The tool supports large-scale conversion through batch processing with retry logic, progress tracking, and parallel execution. Content teams can process large document sets unattended while monitoring conversion progress.

5. Prompt-Driven Formatting Rules

Rather than hardcoding formatting logic, template instructions are stored in editable prompt files. This design allows iCEV’s content team to adjust formatting rules or support future template changes without modifying application code.

6. Azure-Aligned Architecture

The solution integrates with iCEV’s existing Azure ecosystem using:

• Azure OpenAI for language model processing
• Azure Blob Storage for application distribution and auto-updates
• Microsoft Entra ID for secure authentication

This architecture ensured the tool aligned with the company’s existing infrastructure and development practices.

IntelliTect helped us automate a complex content migration that would have otherwise required significant manual effort. Their solution preserved the quality of our instructional materials while giving us a scalable way to modernize our content library.

Steven Lubowicz, Chief Technology Officer, iCEV

Outcome

The resulting system automates the conversion of legacy Publisher and Word documents into standardized Word templates, significantly reducing the manual effort required to modernize iCEV’s content library. Instead of reformatting thousands of documents by hand, the content team can now run batch conversions and focus on reviewing final materials for instructional accuracy.

The tool also provides long-term flexibility. Because formatting instructions are stored in editable prompt files, the system can be reused to reformat documents whenever publishing templates evolve, extending its usefulness beyond the initial migration effort.

By combining practical software engineering with AI-assisted document interpretation, IntelliTect delivered a scalable solution that preserves the integrity of iCEV’s educational materials while enabling efficient modernization of their curriculum library.