On the 11th, Caster <9331> announced that it will use annotation (*) to improve efficiency and quality in EC-related work requested by "CASTER BIZ assistant."

The company fully acquired Grams in June 2024. Grams is conducting annotation for AI development in the EC field, and is verifying its technology using various large-scale language models (LLMs). With Grams joining the company group, the company can incorporate AI technology based on annotation, which assumes continuous accuracy improvement, into EC-related work requested by "CASTER BIZ assistant," enabling rapid completion of work and cost savings to improve customer satisfaction.

At Grams, it provides one-stop service for scratch development or implementation of AI models, annotation tool scratch development. In the future, the company plans to use Grams' technology to expand the services of "CASTER BIZ assistant" through annotation in various industries other than EC.

*: Annotation refers to the act of attaching information called tags or meta tags to data such as text, speech, and images when creating teaching data for machine learning.

*: Teaching data refers to correct answer data describing the correct output to the input. It is used when performing "supervised learning," which is one of the learning methods of AI systems. "Supervised learning" of AI is an algorithm that learns to return the correct output for input.