Document Imaging Conversion

The images are the easy part. The index is the project.

Every imaging conversion looks simple from a distance: the documents are files, copy the files. Then you discover that the files are named by a sequence number, and the only thing connecting them to an account, a document type, a date and a retention rule is an index database in a format the vendor never documented — from a vendor who may no longer exist, or no longer supports the version you are on.

Without the index, you have not migrated a document archive. You have a folder of pictures.

What this work involves

Getting the data out. Sometimes there is an export utility. Often there is not, and the path is reading the index tables directly, matching them to files on disk, and handling the cases where the two disagree — orphaned images, index rows pointing at files that were deleted, duplicates from a failed load years ago.

Understanding the index. Reverse-engineering what each field means, which fields are genuinely used versus abandoned, and how document types map to the new system’s structure. This is a conversation with the people who file the documents, not just a database exercise.

Format conversion. Old systems store images in formats new systems will not accept: single-page TIFFs that need combining into documents, proprietary compression, or images without the metadata modern platforms expect. If long-term retention matters, PDF/A is usually the target.

Mapping to the target. Document types, retention periods, security and folder structure rarely map one-to-one. The new platform is an opportunity to rationalise a taxonomy that grew organically for fifteen years — but only if that is decided deliberately rather than by whatever the script does.

Building the load. Most platforms ingest via an index file plus images in a defined layout. Getting that specification right, and testing it on a small batch before running millions of documents, is what keeps the project from being run twice.

Reconciliation. Document counts, page counts, and spot verification that a document retrieved from the new system is the same document as in the old. This is the evidence that lets you decommission the legacy system — and until you have it, you cannot.

Where this comes up

• The imaging vendor has gone end-of-support, or gone entirely
• The imaging system is tied to a platform being replaced
• An acquisition brings a second imaging system that must be merged
• Records need to move to a modern platform for retention or compliance reasons
• The archive is on hardware nobody wants to keep running

What you get

Extracted images in an accepted format, a rebuilt index mapped to the target’s structure, the load files, the scripts, and reconciliation evidence covering document and page counts.