17. stay with my existing translation management system but make translation memory management easier

You work with a translation management system and feel happy about it. At one point you realize that a year ago, in a certain job, one translator translated a lot of nonsense and your review processes did not spot out the mistakes. A long time has passed, and quite a few segments have been reused - and as they have not been paid for, nobody indicated the mistakes to this day. How do you identify the segments that need a review?

You are right, translation memory tools - no matter whether they are called TMSes or just TMs, whether they are server-based or just local tools - don’t provide you with a toolset to carry out this task. This is why you cannot have a single answer to whether it’s better to have all your translations in a single big translation memory or in several smaller translation memories. Smaller translation memories make you lose leverage, and if you use a large TM, translation memory maintenance becomes a pain. Would you go through each and every segment just to take out 300 segments with mistakes?

When you approach this from the point of view of costs, you will find that you lose money on losing leverage (opportunity costs) and you lose money on TM maintenance. You’re trying to balance between leverage outage and maintenance costs. This is one of the problems that Kilgray’s TM Repository technology can address.

The TM Repository gives you full control over your translation memory units - not just the translation memories. It is a single database with enhanced translation management functionality. The TM Repository is not an online translation memory. It is a cost-saving tool that works best if you have millions of TM entries. It increases leverage and cuts maintenance costs. It can be integrated into any translation management system or used through a web interface. The TM Repository is not part of the memoQ solution and it does not compete with any other commercially available solution. It’s a whole new concept to TM management.

The TM Repository offers the following functionality:

• You can set up a metadata structure that reflects your workflows, product structure and documentation structure.
• It enforces the use of this metadata structure for each and every translation unit and protects the metadata.
• It enables you to create ad hoc translation memories for each and every job by filtering the units in the TM Repository. These ad hoc translation memories are logged with an identifier.
• Once a job is done, it enables you to feed the updated translation memories back into the TM Repository. If the translation tool your vendor used has corrupted the metadata, it restores the metadata, and it archives both the old and new TM entries.
• It provides versioning for your translation memory units.

• Minimize your translation memory maintenance costs and maximize your leverage.
• Flag the translation memory units with all the relevant information and set up translation memories best suited for a given job.
• Protect the translation memory units from corruption.
• Enable rollback operations for all your translation memory maintenance.
• Version your translation memory units.
• Go back to the translation memory of any job you performed in the past. Identify units added throughout the job and identify units added throughout the job that have been updated later. Quickly correct or eliminate all the mistakes.
• Ensure high quality of your translation memories.
• Create new information from existing units. If you have an English to German and an English to Japanese translation, create a German to Japanese unit.
• Allow your vendors to use any translation tool without compromising your leverage.