Rapid information retrieval through trained data and search capabilities
High flexibility by selecting and operating the optimal LLM model for specific business needs
Intuitive UI/UX Design – Easy accessibility and usability for all users
Chatbotchat functionality, conversation history storage, continuous conversation support, file content search, multiple LLM model application
Model Trainingtraining data management, model fine-tuning capability, real-time monitoring of training process
RAG Searchvector database storage, document-based Q&A, advanced search functionality
Document Generationlearning of various document formats, automatic document creation, customized document generation support
HR Manager | Budget Manager | Internal Staff |
 |  |  |
Time Series Data-Based
Image Data-Based Defect
Automated Precision Diagnostics Across Equipment
Parameters Using Health Scores
- Preprocessed Data Infrastructure for Seamless AI Application and Custom Model Development
Customizable AI Models for Equipment-Specific Problem Solving
(e.g., Plasma Chamber Gas Leak Detection)
Modular Architecture for Scalable Integration of
Tailored AI Solutions
01. Data Collection & Preprocessing
Equipment parameter data from the CTC (Cluster Tool Controller) is collected and stored in the source database, then fetched, validated and preprocessed for use in deep learning training and inference.
02.Learning Differences Between Normal and
Abnormal States
A reference is created using 100 normal wafers with acceptable yield levels and the system learns the precise differences between normal and abnormal states by comparing this reference with data from abnormal equipment behavior.
Using a finely trained deep learning model, anomalies are identified by comparing wafer data against normal patterns, instantly providing health scores that indicate the severity and location of abnormalities, along with potential causes in the process or equipment.04. Equipment Anomaly Detection & Process Result Prediction
By thoroughly analyzing equipment parameter data for each wafer, the system detects issues such as plasma chamber gas leaks or equipment failures and predicts process outcomes like the etch amount in etching equipment.
LED X-ray Inspection
LED Visual Inspection
Heating Coil
Defect Inspection
PCB Defect Inspection
Furniture Defect
Inspection
Cosmetic Container
Visual Inspection
Battery Manufacturing
Video Anomaly Detection
Valve Cylinder
Visual Inspection
Time Series Data-BasedProcess Anomaly Detection
Equipment Failure Detection
Health Scoring
Process Outcome Prediction
Root Cause Analysis of Anomalies
Standardization of Collected Data
Image Data-Based DefectDiverse Models (OD, CLA, KD, SEG)
One-Click Labeling
Labeling Recommendations
Insight Generation for Model Performance Improvement
Measurement of Defect Types and Physical Quantities
Model Deployment / Mass Production Integration
Automated Precision Diagnostics Across Equipment
Parameters Using Health Scores
- Preprocessed Data Infrastructure for Seamless AI Application and Custom Model Development
Customizable AI Models for Equipment-Specific Problem Solving
(e.g., Plasma Chamber Gas Leak Detection)
Modular Architecture for Scalable Integration of
Tailored AI Solutions
