*Jidoka – Japanese word that means “automation with human touch”
Data Sources
Customer data sources, Scry AI data sources, data providers
Semi-Structured and unstructured data structure
Time series data
Web source/reference data
Paper/image data
Third party data
Scry AI Platform – Jidoka®
Proprietary library of AI algorithms for data engineering and specific business use cases optimized for compute and memory usage. Ingests & cleanses internal & external semi-structured & unstructured data, harmonizing it for better accuracy & reliability
Proprietary data ingestion, engineering & cleansing platform providing lineage and actionable insights (Jidoka*)
- Digitization library
- Data exploration
- Business rules (BR) engine
- Ontology and live manual
- Connectors & scrapers library
- Role based access & security
- Data quality (DQ) engine
Proprietary library of AI algorithms optimized for computing, power and memory usage that achieve higher accuracy than open source algorithms
- Supervised learning
- Knowledge graph based
- Edit distance based
- Filtering
- Unsupervised anomaly detection
- Unsupervised similarity finding
- Business rules – Probabilistic
- Business rules – Deterministic
- Time series and Natural language processing
- Unsupervised clustering
Products
Built with deep domain expertise achieving unparalleled accuracy of 91% to 98%
Collatio® – Providing accurate, reliable data & its lineage
Annomaly® – Detecting potential fraud, risk & non-compliance
Servis® – Predicting equipment and service failure & its potential causes
For Demo & Additional Information