Most migration solutions can move a document. The difficult question is whether that document arrives the way your organization needs it: correctly classified, with every version intact, open to the correct people and closed to everyone else.
For organizations that handle export-controlled, security-classified or business-critical information, “it arrived” is not good enough. This is why we built a migration tool that goes beyond the basic requirement list, and is making content smarter on the way in.
Why doing more than the basic requirements matters
A migration is a rare moment: every object in the repository passes through one pipeline, once. That makes it the cheapest point in a system’s life to fix metadata, apply classification rules, and leave behind what should not move.
A tool that only copies files wastes that opportunity. The gaps in the old system are carried straight into the new one, and someone pays to fix them later, document by document.
Built for OpenText Content Server
The Turnpikes Migration Tool was designed around OpenText Content Server and Content Management by OpenText (xECM) from the start. It works through the Content Server REST APIs, authenticates via OTDS, and understands the platform’s own object model instead of treating everything as a file with a few fields.
That means business workspaces created from their templates, with folder trees, roles, and permissions. It means compound documents rebuilt release by release, complete version histories, and generations, shortcuts, URLs and task lists re-pointed to their migrated targets.
Every migrated node also carries a permanent reference to its source object, so any item can be traced back to where it came from, years after the migration is finished.
What a standard migration tool does
Most migration tools follow the same pattern: read a list of objects, copy the content and a set of metadata fields, create them in the target and write a log. For a simple file share, that is often enough.
Enterprise content is rarely that simple. Category structures differ between source and target, attribute values have changed meaning over the years, and some content must never leave its current home at all.
What is on the market
Broadly, there are three options. The first is the platform’s own import utilities, such as OpenText Object Importer, which loads objects into Content Server from XML control files and a file system. They load reliably, but translating the old system into those files is left to the project.
The second is general-purpose migration platforms that connect many systems. Their breadth is their focus, but platform-specific structures such as business workspaces, compound documents, and their permission models are often only partly covered, leaving vulnerability or missed opportunities for cleanup.
The third is custom scripts written for a single project. They fit the first job exactly and are then rewritten from scratch for the next one, with every lesson learned along the way. This is a time-intensive problem which requires a large amount of manual effort.
What we have been asked to do on top
Our clients in Defense and Aerospace, Pharma, Logistics, Transport, HR and Financial Services have asked for far more than a copy. In practice, that has meant:
- Security filtering: objects with restricted classifications are held back by rule and recorded as skipped, never silently dropped.
- Export control and data privacy: dependent attributes are filled in automatically whenever a classification requires them.
- Metadata translation: every source attribute is mapped to its target equivalent, with fallbacks, defaults and value translation.
- Live-system migration: delta passes keep the old and new systems in step until cut-over.
- Locked-down environments: a fully offline installation on secured Windows servers, with no credentials stored in code.
- Proof: a status, target ID and error message for every object, plus rehearsal runs before anything touches production.
Where Knowledge Discovery comes in
OpenText Knowledge Discovery (formerly IDOL) is OpenText’s AI platform for understanding unstructured content. It extracts text from more than 2,300 file types, reads scanned material with OCR, recognizes entities and sensitive data such as personal information, and classifies content by meaning.
Recent releases add LLM-based classification that needs no training data, and grammars that recognize export-control terminology, including the ITAR munitions list. Those are exactly the questions a defense migration has to answer for every document.
In the Migration Tool, the transform stage is where each object is already inspected, mapped, and checked against business rules. The tool is designed to call an enrichment service at that point, so Knowledge Discovery can work on every document as it moves from source to target.
That opens three opportunities:
- Before the move: analyze the legacy repository in place to find duplicates, obsolete content and sensitive material, and scope the migration accordingly.
- During the move: run OCR on scanned documents, extract entities and suggest classifications, then write the results into Content Server categories as each object is created.
- After the move: confirm that everything flagged as sensitive arrived with the right classification and permissions.
The result is content that is not just in the new system, but searchable, classified and ready for AI-powered search and assistants from day one.
Why our tool goes further
It is reusable. One engine has been refined over five generations and many client projects, so each migration starts from proven, production-tested code rather than a blank page.
It is resilient. An interrupted run resumes where it stopped; parents are always created before their children, and failed items can be retried on their own without repeating what already succeeded.
It is evolving. The newest generation adds a web client, scans the source system by itself and runs as a supervised Windows service, with named mapping profiles so different migrations can run side by side under their own rules.
Is It Scalable?
We have migrated projects from hundreds of thousands to hundreds of millions of documents and more. Our solution excels at scale because it is built with that in mind. We evaluated this during our implementation, for example, the modular design of the architecture handles each step of the migration in an independent subsystem. This is easy for us to modify and upgrade, keeping our tool simple means that we were able to offer a solution that could easily take on large migration projects with confidence, speed and accuracy.
The tool can be used across industry and in regulated spaces. We have deployed it in Public Sector, Insurance, Financial services and many more industries which are subject to regulation and compliance processes with great success.
What it connects to
Sources: OpenText Content Server and Extended ECM, Documentum, SharePoint Online and on-premise, Salesforce, and file systems or SMB shares.
Targets: Content Server and Extended ECM, and SharePoint Online.
Around the migration: SQL Server staging databases or Excel control sheets to define the scope, Azure Service Bus and Event Hub for downstream events, and a CMIS-ready design. Knowledge Discovery adds the intelligence layer on top.
Conclusion
Most migration tools can move a document. What matters is whether it arrives classified, controlled and traceable, and whether the organization is better off afterwards than it was before.
That is the standard our clients hold us to, and why the Turnpikes Migration solution goes beyond the basic requirement list: preserving what matters, enforcing the rules that protect it and, with Knowledge Discovery in the loop, making content smarter on the way in.
Planning a Content Server migration, or wondering what your legacy repository really contains? We would be glad to talk.
