.png)
In the ICT industry, optimizing repetitive tasks is a key challenge
that determines a company’s responsiveness and productivity.
LG U+ partnered with Letsur to automate government project discovery
and proposal writing through AI, establishing an execution framework
to support the process.
LG U+ is a leading provider of ICT solutions for various business customers (B2B) based on communication services. As diversified as the business portfolio was diversified, similar tasks such as searching for government issues and preparing proposals tailored to companies were being carried out repeatedly internally. However, as this process was not standardized and was carried out manually, there was a problem that significant resources were consumed inefficiently. This was a potential business risk that hindered the company's market response speed and productivity.
To solve this problem, LG U+ went beyond introducing short-term tools and aimed to fundamentally improve its constitution. We wanted to transform a manual work system into an AI-based automation system and build this success story as a 'repeatable execution framework' that can be applied to solving other problems within the organization. The core goals of the project were to “define work with data,” “design a structure where AI can intervene,” and “implement a system that reliably utilizes the results throughout the organization.”
Rather than simply developing technology, Letsur redefined this project as a transformation challenge that fundamentally digitizes the way LG U+ organizes work. We participated as a strategic partner throughout the entire process, from problem structuring to execution framework design to system advancement. This is closely linked to LG U+'s medium- to long-term digital transformation (DX) strategy. This is because it is the first button to create a virtuous cycle structure that maximizes internal operational efficiency, concentrates secured resources on core businesses, and furthermore develops internalized AI capabilities into a new business model.
The first button of the project was to obtain “structured data” for AI to learn and make judgments. To improve the existing manual information gathering process, we've developed a dedicated web crawler that automatically traverses key announcement sites. The collected announcement data was automatically classified and refined according to criteria such as industry sectors, governing bodies, and technical keywords, and integrated into a database that anyone can easily access and search. The proposal automation project also reorganized product information held by LG U+ according to a structured template such as product name, core functions, and application examples. This was a key process that turned subjective and sporadic information into objective 'materials' that AI could understand and use.
The next challenge was for AI to create “criteria for judgment.” An objective evaluation system was needed to select opportunities of real value for LG U+ from numerous government task announcements. Letsur designed an 'Evaluation Scoring' logic that calculates suitability by combining text similarity, project budget, and participation conditions between the project announcement and LG U+ technology. Through this model, it is now possible to evaluate and prioritize all tasks based on the same standards, moving away from the method of relying on the subjective judgment of the person in charge. This scoring system is designed with a flexible structure that can be continuously learned and refined through future accumulated data.
An AI model must have a circular structure that rapidly improves through feedback from business users. To this end, we have built an 'AI Page' interface where users can directly test the developed AI functions and check the results immediately. Through this page, users were able to search for assignments, request the creation of draft proposals, and receive results in real time. The important thing is that beyond simply using the function, we have integrated a feedback loop within the system where you can evaluate your satisfaction with the results or submit suggestions for improvements right away. Through this, the development team was able to identify what features were useful and what level of accuracy the model was with quantitative data and quickly reflect improvements.
The final challenge was to create an environment where the developed AI functions could be used stably throughout the organization. Letsur configured a dedicated LG U+ workspace within its own platform 'Staix', and linked the system so that these functions are provided only to internal users in a strong security environment. More than a simple function delivery window, this workspace includes features essential for AI governance, such as managing AI model versions, tracking usage records for each user, and managing feedback history. This increased organizational trust by transparently managing who used AI functions, when, and for what purpose, and laid a solid foundation for expanding AI applications in the future.
The most visible result is a dramatic reduction in working hours. Government task search tasks have been replaced by crawler-based automated dashboards, and the time spent on related tasks has been reduced by approximately 80% compared to the previous one. This means that managers have moved away from mechanical information collection to focus on high-value tasks of analyzing collected information and formulating business strategies. The automated proposal system was implemented so that only key elements are entered and a highly finished PPT draft is generated within minutes. As document creation and design tasks that used to take hours have been reduced to minutes, the speed of responding to customer requests has improved dramatically, which directly contributes to securing business opportunities.
This project was an opportunity for LG U+ to shift its decision-making method from subjective experience to an objective data base. In particular, the introduction of an 'Evaluation Scoring' system to evaluate suitability for government tasks has unified judgment criteria that were different for each department and person in charge across the enterprise. As all members discussed whether to participate in the task based on the same criteria and quantified scores, unnecessary disagreement coordination time was reduced and the speed and quality of decision-making improved simultaneously. This changed the way of working that depended on the competencies of specific individuals to one supported by systems and data, and had the effect of up-leveling the ability of the entire organization to perform tasks.
The biggest strategic achievement of this project is not only the development of a one-off tool, but also the successful internalization of a circular structure, or “execution framework,” that “experiments, verifies, improves, and utilizes” AI functions within the organization. In addition to this challenge, this framework will function as a core infrastructure in the process of solving another internal inefficiency problem that LG U+ will face in the future with AI. Furthermore, LG U+ is capitalizing this example of internal innovation into its competitive solution and transforming it into a product form that can be proposed to external corporate customers. This is a typical success story of establishing an ideal virtuous cycle where investment in internal efficiency develops into a business model that generates new profits.
The LG U+ project shows that 'how to define a problem and design an action plan' is far more important than the technology itself for successful AI adoption.
First, the project focused on building a 'framework' that fosters the organization's problem-solving capabilities rather than individual automation tools. The series of processes of problem definition, data collection, quantification of judgment criteria, functional verification, and governance integration will function as a standard process for solving similar problems that LG U+ will face in the future. It created a repeatable system of success rather than a one-time success.
Second, expansion strategies that link internal innovation to external commercialization present new possibilities for AI projects. LG U+ is developing a system developed to improve internal work efficiency as a showcase that demonstrates its DX capabilities and a solution that can be sold to other companies. This suggests that AI adoption should be viewed not from a cost perspective, but from an investment perspective that creates new value.
Lesser's role in this process was not just development. We specifically analyzed the needs based on the customer's business, derived a 'realistically workable design' based on this, and worked together throughout the project as an 'end-to-end partner' responsible for the final development. By combining technical implementation capabilities with strategic ability to structure business problems, we were able to achieve results that fundamentally changed the organization's capabilities.
This project is an innovative example of how LG U+ went beyond simple task automation through AI and redesigned the fundamental framework of how we work. We have completed a virtuous cycle of defining tasks with data, establishing judgment standards with AI, and evolving the system with feedback from the organization. These structural changes will be a key driving force for solving various business challenges that LG U+ will face in the future, and internalized capabilities will be the foundation for new business opportunities that can be expanded externally.
As a strategic partner that goes beyond solving immediate problems and builds capabilities and systems for customers to innovate on their own, Letsur will lead the company's most essential digital transformation.
.svg)
