Model catalogue · Amazon Web Services
Amazon Embedding
Amazon Web Services' embedding model turns files and pleadings into numerical vectors so that a search across the corpus hits on substance rather than wording alone - it finds passages that are close in content even when the words searched for do not appear there. It produces neither text nor assessment. Via Amazon Bedrock it is evidenced region-bound in Frankfurt, without an inference profile that could route onwards, without operator access and without storage in the delivered state. The contractual chain nevertheless remains incomplete: for the confidentiality undertaking under section 203 of the German Criminal Code there is no publicly available, signable document on this route - that element is missing, and with it the coverage.
You can work with it - but part of the safeguard is open: either the provider does not commit to confidentiality itself, or it depends on an agreement you still have to conclude. Which of the two applies to which route is set out in the full matrix.
What it is good for
- Evidenced region-bound in Frankfurt - processing stays in Germany, without an inference profile
- Text and multimodal variants in the same contractual position
- No operator access and no storage in the delivered state
Where it stops
- Embedding is a side role - it does not replace a language model but arranges the corpus for its work
- A US group remains in the chain, even with processing in Germany
- Hit quality depends on how the file is prepared, not on the model alone
Status through us
The same model, procured differently, is a different legal position. The status above refers to procurement through us; processing takes place in a third country.
Which route carries it, with which instrument and under which condition, we set out in full to the firm before it commissions us - section 203(4) requires that the chain can be checked, not that it is published. The full matrix, with the sources behind every route, can be requested in the German catalogue after confirming your e-mail address, and we go through it in the conversation.
Is Amazon Embedding the right fit?
We go through your tasks and the route that carries them - and say plainly where it does not.
Arrange a conversation