SK C&C "Opening the Era of Customized 'Small Giant Language Model DIY' for Enterprises"

Introducing 'Solue LLMOps', the sLLM Implementation Platform

An era is dawning where companies can self-implement customized artificial intelligence (AI) services using small large language models (sLLM) that freely combine various foundation models.

On the 20th, SK C&C announced that it will provide 'Solur LLMOps,' a platform that supports companies in implementing customized small large language models (sLLM).

Solua LLM Ops launched by SK C&C
Photo by SK C&C

Solur LLMOps first supports the combination and utilization of various foundation models. It can utilize commercial LLMs such as OpenAI's ChatGPT and Naver Cloud's HyperCLOVA X, as well as models created with open-source LLMs. It supports the process of recommending, selecting, combining, and utilizing the models necessary to build sLLMs tailored to the characteristics of each company.

In particular, hyper-automation is applied throughout the entire process tailored to corporate tasks, including ▲ data collection and preprocessing ▲ automatic training data generation ▲ training using external foundation models ▲ sLLM creation and testing. It also provides a function to eliminate hallucination phenomena, where AI generates incorrect answers.

Additionally, it offers a user-friendly environment (UI) and experience (UX). Users can select data on a web screen and repeatedly perform tasks from data refinement to model tuning and testing through a simple drag & drop method. It is easy to handle large amounts of data and perform simultaneous parallel training using multiple foundation models with just mouse operations. This enables practitioners in the field to easily create sLLMs necessary for their own work.

Jiwon Cha, Head of the G.AI Group at SK C&C, said, "The main functions contained in Solur LLMOps have been applied to generative AI construction projects at SK affiliates, including financial companies," adding, "Starting with this, we will actively strive to expand customized sLLMs for companies across the entire domestic industry."