An AI development platform
that can easily derive prediction results using data you own.

Even if you are not a data scientist, you can easily develop an AI model and apply it to business
without the burden of professional manpower and cost for AI project promotion.

기업이 AI 구현에 겪는 어려움. 인프라부족 35%, AI인력부재 18%, 비용효율 18%, 부정적영향 25%, 기술검증우려 25%
In order to introduce AI

In order to advance an AI project, you need an expert who can perform the processes of data preprocessing, model training, and prediction of results.

However, in most general enterprises, there are difficulties in terms of professional personnel and cost to drive the AI process.

1단계 데이터 전처리 : 데이터 준비 및 공학적인 데이터 변환, 2단계 모델 학습 : 알고리즘 제공 및 AI 학습 진행, 3단계 모델 실행 : 직관적 모델 평가 지표 시각화 기반 결과 화면, 4단계 모델 운영 : 주기적인 모델 모니터링 및 진화 관리
Applying AutoML to solve these problems

Automated Machine Learning (AutoML) enables non-professionals to develop AI models and apply them in the field without a separate coding process. Even ordinary users can easily work with AI with just a few mouse clicks.

Application of artificial intelligence without separate coding
Utilization of development tools by non-development personnel
Reduce AI manpower and related costs
Applicable to various industries

Currently, by providing industry-specific AI applications with predictive maintenance, abnormal transaction detection, and personalized recommendations,
It has been applied to various industries such as manufacturing, energy, education, environment, media, distribution and public sector.
In addition, our technological prowess is being recognized by acquiring first grade GS certification in 2020,
selecting an innovative prototype by the Public Procurement Service in 2019, and listing in the Gartner newsletter.

good software logoAcquired 1st grade GS certification
조달청 로고Selected as an innovative product by the Public Procurement Service

System configuration diagram

  • Data maintenance
    데이터 전처리:데이터탐색 데이터변환 특징선정, 모델학습:모델구축 모델평가 모델최적화, 예측:모델적용 모델모니터링 시각화
  • Algorithm
    클러스터링, 분류, 회귀, 딥러닝:시계열예측 이미지분류
  • Collection and Storage
    Data Lake
  • Source Data
    정형, 비정형, 이미지, 센서, 로그

WiseProphet™ main features

  • 01
    Data preparation
    • Support various structured and unstructured data
  • 02
    Engineering data preprocessing
    • Data transformation, data cleansing, scaling
    • Feature extraction by text vectorization
    • Analysis of importance and correlation for feature selection
  • 03
    Support various algorithms
    • Support for classification, regression, and deep learning algorithms
    • Support for unsupervised learning clustering
    • Support for new algorithms (Xgboost, LSTM)
  • 04
    Automatic model management
    • Easy deployment of models to various environments
    • Periodic model monitoring and evolution management
  • 05
    Explain model visualization
    • Provide visualization of model prediction results
    • Provided as a dashboard for intuitive understanding of forecasting results
  • 06
    Algorithm optimization
    • Iterative model training
    • Hyperparameter tuning
  • 07
    Learning machine learning using tutorials
    • learn how to learn data
    • Learn step-by-step how to create an artificial intelligence model using the tutorial function

WiseProphet™ industry-specific applications

Predictive maintenance Predict the maintenance time of equipment and the demand for replacement of parts to increase equipment operability and reduce maintenance costs
personalized learning Recommendation of the best customized curriculum for each individual
Preclinical analysis Recommendation of the best customized curriculum for each individual
smart water treatment Optimization of sewage treatment process by analyzing water treatment process operation data and water quality measurement data
energy efficiency Optimize energy consumption through autonomous and integrated factory energy management using AI technology
Abnormal Transaction Detection Detection of abnormal transactions such as unfair claims, illegal supply and demand, and unfair transactions
위세아이텍 AI Reference