World's first generative AI appliance created by Saltlux
Professional developers and high costs are required in order to apply customized generative AI solutions within enterprises. However, LUXIA-ON is equipped with Saltlux’s LLM, LUXIA, by default, with anyone able to introduce this generative AI at an innovative cost, and it can be used immediately by simply turning on the power. Difference in technologies makes a difference in solutions and innovates customers’ experiences.
- #generative AI appliance
- #LUXIA GPT
- #LLM
- #RAG
- #no code
- #low introduction costs
- #mounted generative AI


LUXIA-ON's distinction
The world’s first generative AI appliance equipped with LUXIA, Saltlux’s LLM, by default. Experience generative AI solutions immediately without an AI professional by simply turning on the appliance.
-
Point 01
High performance and assured stability
Quick inference and redundant structure
-
Point 02
Support of fine tuning
Custom LLM created on your own
-
Point 03
Various APIs
Development linkage and various use cases
-
Point 04
Vector search and RAG
Overcoming LLM limitations and improving quality
-
Point 05
Web-based management tools
Can be easily managed even by non-professionals
-
Point 06
Unlimited expansion support
Easily responds to an increase in the number of users
Main functions of LUXIA-ON
World's first LLM-equipped generative AI appliance, LUXIA-ONNow, generative AI can be used immediately by simply turning it on.
-
Chats with LUXIA GPT
-
Conversations with and researching uploaded documents
-
RAG using vector search
-
Slave support for OpenAI and Open LLM models
-
Provides API sets to support functionality expansion and development
-
Real-time monitoring and web-based management tools
Core Technology
Generative AI appliance LUXIA-ON
World’s first generative AI appliance, LUXIA-ON
Hardware information
Tool Introduction
-
01LUXIA CHAT
-
Smart LLM - Chats with LUXIA
- Provides an advanced chatting interface
- Q&A, generation of different sentences, business support and common sense
- Document (upload)-based Q&A
- Linkage with file management systems
-
-
-
02RAG support
-
Overcoming LLM Constraints with RAG
- Retrieval Augmented Generation
- Processes refer to a trustworthy knowledge base
- Extracts knowledge from the classification of documents uploaded by users and links LLM answers
- Manages documents in linkages to the file management system
-
-
-
03Support of custom embedding
-
Various options are provided for efficient embedding
- Custom embedding models are available.
- Number of chunks can be adjusted.
- Overlapping can be adjusted.
- Optimization embedding consulting is provided after document structure analysis (at a cost).
-
-
-
04Support of various API sets
-
Support for unlimited functionality expansion through API
- Provides a set of APIs that handle the entirety of functions of the appliance
- Indexing (separate/whole), lists
- Search (keyword/vector/ensemble)
- Open AI API linkage
- LLM settings hyper-parameters
-
-
-
05Web-based management tools
-
Real-time monitoring and easy management
- Various management systems (permissions) can be established
- System status monitoring and warning
- Real-time GPU monitoring
- File management system installed
-
-
-
06Support for unlimited expansion (Scale Out)
-
Manages scale by adding as many appliances as desired
- Flexibly responds to an increase in the number of users or storage additions
- Scale out with just simple settings
-
-
-
07Professional consulting and development support (charged)
-
Possible to respond to specialized models or use cases
- Creates a business-specific model
- Professional fine tuning prepared by a prompt engineering team
- Prior learning/fine tuning are possible without GPU servers
-
-
Use Case
Innovate your way of working easily with generative AI appliance LUXIA-ON.