Introduction to technologies

• 01Utilization of various large language models (LLM)

  • Utilization of various large language models (LLM)

    1. LUXIA, Saltlux’s LLM, as well as various other LLMs, which can be utilized through the cloud and
       the MLOps environment, are provided.
    2. These LLMs can be used selectively in accordance with the purpose of service in order to
       construct a flexible and expandable system.
  • 

• 02Model learning through large-scale data processing

  • Model learning through large-scale data processing

    1. Unlike on-premises, that have high initial investment costs for model learning and limited expandability,
       the cloud, which provides flexible computing power and storage, is ideal for AI model learning
       or data analysis using large datasets.
    2. A learning environment optimized for a cloud environment can be constructed and
       a customized model can be created.
  • 

• 03Stable response to rapid traffic variation

  • Stable response to rapid traffic variation

    1. Since resource expansion and automatic scaling are possible in the cloud,
       it is possible to flexibly respond to rapid changes within service demands.
    2. Compared to on-premise environments, the services provide operating stability while preventing resource wastage at peak times.
  • 

• 04Data linkage for RAG (Retrieval-Augmented Generation)

  • Data linkage for RAG (Retrieval-Augmented Generation)

    1. When it is intended to link internal enterprise data with external data collected in real time and
       create a distributed architecture for a cloud through an LLM, data can be efficiently linked and
       managed without a delay by using a cloud.
  • 

• 05Highly available services and distribution automation

  • Highly available services and distribution automation

    1. For services that require operation 24/7, or are intended for general users,
       effective operation management is possible by utilizing highly available and automated CI/CD systems
       embedded in the cloud.
    2. Cross regional distribution can also be useful when targeting global services.
  • 

• 06Collaboration and co-work

  • Collaboration and co-work

    1. For services involving multiple businesses and institutions, cloud-based collaboration tools are convenient.
    2. It is easy to share data and resources and provides excellent accessibility.
  •