Veloc, sLLM-Equipped Proprietary Framework Advancement... "Solution Development Reduced by 30%"

Veloc has recently succeeded in advancing its proprietary solution development framework, ‘B-frame,’ launched to apply the ‘lightweight large language model (sLLM).’ Veloc announced on the 8th that it has independently developed new solutions based on B-frame and completed performance testing.

Veloc launched B-frame in July to maximize solution development efficiency. B-frame’s advantages lie in significantly shortening the solution development period through various automation features such as △dynamic menu generation △automatic HTML generation △template engine integration △automatic conversion between models.

During the development of a new customer management solution using B-frame, Veloc reduced the production time by more than 30% compared to before and confirmed through testing that a high-performance solution was implemented. Since B-frame-based solutions allow integrated management of solutions within a single framework, it will also be easy to equip generative artificial intelligence (AI) functions such as sLLM in the future.

sLLM is characterized by being a lightweight version of LLM (large language model) optimized for its intended use to maximize learning speed. Veloc plans to equip sLLM in B-frame within this year to add generative AI functions to all solutions. With the addition of generative AI functions, if the framework learns its own database autonomously, it is expected that the speed of enhancing Veloc’s solution performance will be maximized.

A Veloc representative emphasized, “Through B-frame, not only can the production time and cost required for new solution development be significantly reduced, but integrated management also enables faster response to customer requirements. Based on differentiated technological competitiveness, we will actively secure new customers and expand sales and profits.”

He added, “We are sequentially conducting integrated work based on B-frame for existing products and are further advancing functions aiming to equip sLLM within B-frame by the end of the year. When generative AI functions are applied to all solutions, not only development speed but also performance can be maximized, so we expect to leap to a leading position in the market.”