Korea Elevator Safety Agency Educational Administration Support AI Chatbot
KoELSA Educational Administration Support AI Chatbot
OWLIX AI Content - Korea Elevator Safety Agency (KoELSA) Educational Administration Support AI Chatbot
The complex educational administration system and technician career reporting procedures of the Korea Elevator Safety Agency have caused a high workload for both civil petitioners and administrators. To resolve this administrative inefficiency, Samwoo Immersion has built an 'Educational Administration Support AI Chatbot' that combines OWLIX AI's LLM (Large Language Model) with a sophisticated Rule Engine. Going beyond simple answers, this solution provides an innovative administrative environment that analyzes national public data APIs in real-time to map customized educational courses by building and allows the AI to self-verify qualification requirements by reading complex legal evidentiary documents.
Speed
Accuracy
Security
Efficiency
Information acquisition time reduced by over 90% compared to existing phone/visit inquiries
Achieved analysis accuracy of over 98% based on agency management standards
Perfect privacy protection through non-storage and immediate deletion structure
30% reduction in document rejection rate and maximized efficiency via intelligent self-verification
INTRO
Lack of Information Accessibility ยท Verification Overloadย ยท Legal Compliance Errors ยท Data Management Burden
"Diagnosing the factors hindering administrative efficiency: the cycle of phone inquiries and document rejections" In KoELSA's educational administration, the mandatory legal education required varies by the building's height, purpose, and area, making it very difficult for petitioners to accurately identify the correct information themselves. Furthermore, the verification of numerous evidentiary documents during the elevator technician career reporting process creates a heavy workload for administrators, leading to persistent dissatisfaction among petitioners due to frequent document rejections.
SOLUTION:ย ย "One-stop Administrative Innovation Realized through Public Data Integration and LLM Technology" This is a customized system that automates KoELSA's complex educational system and technician career reporting procedures using the latest AI technology. Beyond a simple chatbot, it features building analysis via real-time API integration with VWorld and the Ministry of Land, Infrastructure and Transport (MOLIT), and a document verification engine based on an LLM that has perfectly learned Article 28 of the Enforcement Decree of the Elevator Safety Management Act. Through this, information accessibility for petitioners is dramatically improved and the administrative workload for managers is fundamentally resolved.
1.ย Real-time Data Integration
Ensured data accuracy by querying nationwide building information in real-time through the MOLIT API.
2.ย AI Auto-Reading
Artificial intelligence determines complex legal qualification requirements with over 98% accuracy through an LLM-based career self-verification system.
3.ย Privacy Protection
Designed with a volatile structure that deletes documents immediately after analysis without storing them, fundamentally blocking security incidents.
ย ย
Configuration of KoELSA Educational Administration Support OWLIX AI Chatbot
01. Real-time Education Guidance Based on Address When a user enters a road name address or building name, the system queries the building ledger in real-time. It accurately guides the user to the legally mapped course for that building, such as 'Evacuation Elevator Safety Manager Training', 'Emergency Rescue Operation Training', or 'General Elevator Safety Manager Training'.
02. AI-based Career Reporting Self-verification When up to 4 types of required documentsโsuch as Health Insurance Eligibility Certificate, Employment Certificate, Graduation Certificate, and Qualification Certificateโare uploaded, the AI analyzes real-time whether legal qualification requirements are met.
03. Intelligent Supplementation Guide Generation The AI precisely determines if qualification requirements are met according to Article 28 (Recognition Standards by Technician Grade) of the Enforcement Decree of the Elevator Safety Management Act. If requirements are insufficient, it generates customized text such as "You are 9 months short of practical experience after obtaining the Elevator Technician qualification," providing specific reasons for the deficiency and required additional evidence.
04. Dual UI (Screen Styles) to Suit Your Preferenceย
Professional Mode: Configured to allow users to grasp professional information at a glance, without additional elements like characters. Character Mode: Uses the cute OWLIX character and soft pastel colors to guide users through difficult administrative tasks in a friendly manner.
Professional Mode
Character Mode
Specialized Technologies for OWLIX AI-based Intelligent Education Administration and Qualification Verification AI Chatbot
โWe have implemented advanced administrative intelligence through Samwoo Immersionโs unique architecture and sophisticated algorithms.โ
1. Technical Differentiation and Strengths of OWLIX AIย "An intelligent sLLM-based learning engine that builds trust under controlled information"
ย Category
ย Conventional AI
ย OWLIX AI (sLM based)
ย Training Data
ย Random learning of massive global data
ย Selective learning only from verified information and ย directly input data
ย Information ย Reliability
ย Low reliability due to hallucinations
ย Maximized reliability based on curated data
ย Learning ย Support
ย Provides information not tailored to ย individual user needs
ย Customized learning support within user-required ย ย ย scopes
ย Operational ย Efficiency
ย Causes information confusion and ย unnecessary data overload ย ย ย
ย Strengthens learning focus and efficiency by ย providing filtered info
ย Security
ย Security vulnerability concerns due to ย external cloud reliance
ย Maintains strong data security through on- ย ย premise/volatile processing
1. Customized sLLM Technologyย (smaller Large Language Model)ย
Unlike existing LLMs that learn random data, it learns only verified information within the range defined and entered directly by the instructor or user, delivering accurate information without hallucinations.
2.ย Intelligent Rubric Evaluation System
Through multi-dimensional evaluation functions where AI automatically scores and analyzes the learner's performance, it provides quantified metrics for problem understanding, logic, and accuracy.
3.ย AI Virtual Character to Increase Brand Familiarity
By introducing 'OWLIX', an owl character symbolizing wisdom and exploration, we transform rigid administrative services into a friendly communication point and maximize learning immersion.
4.ย Multi-Interface and Device Support
Provides consistent AI interaction experiences across various Extended Reality (XR) environments such as VR, AR, and Metaverse, as well as web and mobile.
1-1. OWLIX AI System Architecture (System Diagram) "Next-generation intelligent architecture through the innovative convergence of AI and XR"
1.ย Input Layerย
User voice (STT) and text data, career evidentiary document uploads
2.ย Analysis Layer
sLM Engine: Natural language understanding and response generation based on administrative laws Logic Engine: Rubric-based multidimensional evaluation (analysis of problem understanding, logic, and accuracy)
3.ย Data Layer
MOLIT real-time building data API and dedicated administrative Knowledge Base
4.ย Output Layer
ย Issuance of AI-generated comprehensive feedback reports and updates to visualization dashboards
2. Technical Differentiation and Strengths of Educational Support AI Chatbot "Intelligent civil service solver completed with public data integration and precise judgment logic"
ย Category
ย Existing Educational AI Services
ย Educational Administration Support AI Chatbot
ย Information ย Delivery
ย Focused on general educational info or ย terminology
ย Real-time building-education mapping based on ย address
ย Verification ย Function
ย Users must manually compare and verify ย documents
ย Real-time career/qualification self-verification via AI ย OCR ย
ย Feedback ย Level
ย Answer checking or simple result ย notification
ย Detailed supplementation guide (e.g., "9 months ย short of experience")
ย Legal Basis ย
ย Answers based on general knowledge ย ย ย
ย Precise judgment based on latest legal logic like ย Safety Management Act
2-1. Intuitive User-Centered Service Process When accessing the system, users can intuitively select the administrative service they need without exploring complex menus. We have established a process that allows access to two core solutionsโ'Address-based Guidance' and 'AI Career Verification'โwith a single click, without complex menu navigation.
User-centered intuitive service process
Service selection screen
2-2. AI Integrated Verification Engine (Process from OCR to Precise Judgment) This integrates the 'recognition' stage, which reads information from documents, and the 'inference' stage, which interprets it according to laws, into one organic process. First, through OCRService (Optical Character Recognition), characters in documents like Health Insurance Eligibility Certificates or Employment Certificates are accurately read as digital data. Then, an LLM (Large Language Model) grasps the context to organize the scrambled text into an analyzable format (data structuring), such as career period, type of qualification, and major status. This refined data is immediately passed to the Rule Engine (Intelligent Judgment System) to be precisely analyzed according to Article 28 of the Enforcement Decree of the Elevator Safety Management Act. In particular, the AI calculates and determines recognition even for tricky legal conditions such as [holding a specific qualification + mandatory practical career experience], as in the case of an 'Elevator Industrial Engineer'. This entire process is completed within high-speed processing of about 10 to 30 seconds.
Educational course classification analysis
Technician career processing reference data
AI Integrated Verification Engine Intelligent Service Process
AI analysis screen
2-3. Logic Tree (Intelligent Classification Algorithm) Based on more than 23,000 nationwide building records, the system analyzes complex variables such as skyscrapers with 30+ floors (Case 1), multi-use facilities with 16+ floors (Case 2), and general buildings (Case 3) to precisely map educational courses into 4 cases. In particular, by applying multi-branch logic that considers even special cases such as old buildings completed before 2017 or those with ambiguous floor and height standards, we have implemented an intelligent guidance system where users are immediately connected to mandatory legal education courses just by entering an address.
Design logic
Elevator education target building classification system
2-4. Fallback (Data Correction System) We applied powerful supplementary technology to ensure service continuity even for old buildings (approx. 15%) where national public API information is missing or address systems do not match. Through PNU (Parcel Number) transformation processing and regular expression-based preprocessing logic, typos or omissions in address text are corrected in real-time. This has raised the information provision success rate to over 95%, perfectly filling gaps in administrative services.
Item
Verification Data Count
Successes
Accuracy
Note
Auto Classification of Building Info
23,890
23,890 cases
98.2%
Total of 4 Cases (Evacuation/Multi-use, etc.)
API Fallback Info Provision
3,584 cases
3,405 cases
95%
Estimation logic applied based on floor count when data is missing
AI Career Document Analysis
50 per type
245 cases
98%
Pre-blocking logic for low-quality images applied
2-5. Hybrid AI Engine (Intelligent Infrastructure Realizing Security and High Performance Simultaneously)
To satisfy the strict security requirements of public institutions while utilizing the powerful performance of modern AI, we designed a 'Hybrid Inference Engine' architecture. By combining internal proprietary AI within the security network with the latest external AI, we achieved both security and performance.
Technology method
Supported Language
Usage Environment
sLM-based intelligent learning engine
Korean
Web browser-based service
ย Recommendedย ย Specsย
Built as a web-based service, users can access it without separate installations, provided they have an internet connection and a web browser.
Case Study: KoELSA Educational Administration Support AI Chatbot
The moment you immerse, you grow!ย Maximize your capabilities with XR and AI (X-RAI) technology.
If you have any questions about this content or purchase-related inquiries fromย Samwoo Immersion, please click the 'Inquiry' button below for more detailed guidance.ย
KoELSA Educational Administration Support AI Chatbot
OWLIX AI Content - Korea Elevator Safety Agency (KoELSA) Educational Administration Support AI Chatbot
The complex educational administration system and technician career reporting procedures of the Korea Elevator Safety Agency have caused a high workload for both civil petitioners and administrators. To resolve this administrative inefficiency, Samwoo Immersion has built an 'Educational Administration Support AI Chatbot' that combines OWLIX AI's LLM (Large Language Model) with a sophisticated Rule Engine. Going beyond simple answers, this solution provides an innovative administrative environment that analyzes national public data APIs in real-time to map customized educational courses by building and allows the AI to self-verify qualification requirements by reading complex legal evidentiary documents.
INTRO
Lack of Information Accessibility ยท Verification Overloadย ยท Legal Compliance Errors
ยท Data Management Burden
"Diagnosing the factors hindering administrative efficiency: the cycle of phone inquiries and document rejections"
In KoELSA's educational administration, the mandatory legal education required varies by the building's height, purpose, and area, making it very difficult for petitioners to accurately identify the correct information themselves. Furthermore, the verification of numerous evidentiary documents during the elevator technician career reporting process creates a heavy workload for administrators, leading to persistent dissatisfaction among petitioners due to frequent document rejections.
SOLUTION:ย
ย
"One-stop Administrative Innovation Realized through Public Data Integration and LLM Technology"
This is a customized system that automates KoELSA's complex educational system and technician career reporting procedures using the latest AI technology. Beyond a simple chatbot, it features building analysis via real-time API integration with VWorld and the Ministry of Land, Infrastructure and Transport (MOLIT), and a document verification engine based on an LLM that has perfectly learned Article 28 of the Enforcement Decree of the Elevator Safety Management Act. Through this, information accessibility for petitioners is dramatically improved and the administrative workload for managers is fundamentally resolved.
ย ย
Configuration of KoELSA Educational Administration Support OWLIX AI Chatbot
01. Real-time Education Guidance Based on Address
When a user enters a road name address or building name, the system queries the building ledger in real-time. It accurately guides the user to the legally mapped course for that building, such as 'Evacuation Elevator Safety Manager Training', 'Emergency Rescue Operation Training', or 'General Elevator Safety Manager Training'.
02. AI-based Career Reporting Self-verification
When up to 4 types of required documentsโsuch as Health Insurance Eligibility Certificate, Employment Certificate, Graduation Certificate, and Qualification Certificateโare uploaded, the AI analyzes real-time whether legal qualification requirements are met.
03. Intelligent Supplementation Guide Generation
The AI precisely determines if qualification requirements are met according to Article 28 (Recognition Standards by Technician Grade) of the Enforcement Decree of the Elevator Safety Management Act. If requirements are insufficient, it generates customized text such as "You are 9 months short of practical experience after obtaining the Elevator Technician qualification," providing specific reasons for the deficiency and required additional evidence.
04. Dual UI (Screen Styles) to Suit Your Preferenceย
Professional Mode: Configured to allow users to grasp professional information at a glance, without additional elements like characters.
Character Mode: Uses the cute OWLIX character and soft pastel colors to guide users through difficult administrative tasks in a friendly manner.
Specialized Technologies for OWLIX AI-based Intelligent Education Administration and Qualification Verification AI Chatbot
โWe have implemented advanced administrative intelligence through Samwoo Immersionโs unique architecture and sophisticated algorithms.โ
1. Technical Differentiation and Strengths of OWLIX AIย
"An intelligent sLLM-based learning engine that builds trust under controlled information"
(smaller Large Language Model)ย
1-1. OWLIX AI System Architecture (System Diagram)
"Next-generation intelligent architecture through the innovative convergence of AI and XR"
sLM Engine: Natural language understanding and response generation based on administrative laws
Logic Engine: Rubric-based multidimensional evaluation (analysis of problem understanding, logic, and accuracy)
2. Technical Differentiation and Strengths of Educational Support AI Chatbot
"Intelligent civil service solver completed with public data integration and precise judgment logic"
2-1. Intuitive User-Centered Service Process
When accessing the system, users can intuitively select the administrative service they need without exploring complex menus. We have established a process that allows access to two core solutionsโ'Address-based Guidance' and 'AI Career Verification'โwith a single click, without complex menu navigation.
2-2. AI Integrated Verification Engine (Process from OCR to Precise Judgment)
This integrates the 'recognition' stage, which reads information from documents, and the 'inference' stage, which interprets it according to laws, into one organic process. First, through OCRService (Optical Character Recognition), characters in documents like Health Insurance Eligibility Certificates or Employment Certificates are accurately read as digital data. Then, an LLM (Large Language Model) grasps the context to organize the scrambled text into an analyzable format (data structuring), such as career period, type of qualification, and major status. This refined data is immediately passed to the Rule Engine (Intelligent Judgment System) to be precisely analyzed according to Article 28 of the Enforcement Decree of the Elevator Safety Management Act. In particular, the AI calculates and determines recognition even for tricky legal conditions such as [holding a specific qualification + mandatory practical career experience], as in the case of an 'Elevator Industrial Engineer'. This entire process is completed within high-speed processing of about 10 to 30 seconds.
2-3. Logic Tree (Intelligent Classification Algorithm)
Based on more than 23,000 nationwide building records, the system analyzes complex variables such as skyscrapers with 30+ floors (Case 1), multi-use facilities with 16+ floors (Case 2), and general buildings (Case 3) to precisely map educational courses into 4 cases. In particular, by applying multi-branch logic that considers even special cases such as old buildings completed before 2017 or those with ambiguous floor and height standards, we have implemented an intelligent guidance system where users are immediately connected to mandatory legal education courses just by entering an address.
2-4. Fallback (Data Correction System)
We applied powerful supplementary technology to ensure service continuity even for old buildings (approx. 15%) where national public API information is missing or address systems do not match. Through PNU (Parcel Number) transformation processing and regular expression-based preprocessing logic, typos or omissions in address text are corrected in real-time. This has raised the information provision success rate to over 95%, perfectly filling gaps in administrative services.
(Evacuation/Multi-use, etc.)
low-quality images applied
2-5. Hybrid AI Engine (Intelligent Infrastructure Realizing Security and High Performance Simultaneously)
ย Specsย
Case Study: KoELSA Educational Administration Support AI Chatbot
The moment you immerse, you grow!ย
Maximize your capabilities with XR and AI (X-RAI) technology.
If you have any questions about this content or purchase-related inquiries fromย
Samwoo Immersion, please click the 'Inquiry' button below for more detailed guidance.ย
Inquiry
๐ Phoneย ใ ฃ ย 1600-1663 ย ย โ Emailย ใ ฃ ย info@samwooim.com
* Please include your company information and contact details.
๐ Hoursย ใ ฃ ย Weekdays 9:30~18:30 ย * Lunch: 12:30 ~ 13:30ย (Closed on Sat, Sun, and Public Holidays)