PROFESSIONAL TECHNOLOGY ASSISTANCE
We make technology accessible!
ABOUT US
The Technology
CogniMarts is an AI-based ETL-automation tool for companies that have multiple data sources and need to gain insights from them quickly. CogniMarts automates the Data Marts modeling by recognizing relationships between tables, and discovering dependencies and schedules to provide timely no-code solutions ready for data consumption. Unlike other ETL automation tools that rely on source-provided metadata, CogniMarts utilizes state-of-the-art Machine Learning algorithms to identify the critical components needed to model, build and productionalize a Data Warehouse. Therefore, instead of wasting valuable human effort in the modeling, deployment, and maintenance of the whole ETL process, CogniMarts automates them and minimizes the needed human effort to 10% of the effort required in traditional endeavors. CogniMarts recognizes data types in the modeling phase, identifies primary keys, recognizes relationships, and suggests source-driven data marts. In addition, it identifies data outliers and performs data cleansing and standardization. Likewise, it facilitates maintain SCD types 0, 1, and 2 automatically. Requirement-driven data marts are also facilitated by offering variously computed and derived data. CogniMarts tracks schema drifting in the production setting, and if allowed, it automatically reflects the new schema of the data in all dependent zones – both dimensions and fact tables. Likewise, it supports complete data lineage visualization and tracking of each record to its origin or destination. Error handling also allows for reprocessing late-arriving data. Different processing orchestration and scheduling are supported – per defined schedule or event-triggered.
KEY SUCCESS FACTOR TO SCALE ANALYTICS
While organizations from all industries have been fighting for years to scale their business analytics capabilities and are now massively adopting Data Lake architectures and Hadoop platforms to try to accelerate the development of their analytic maturity, one basic fundamental obstacle remains: data preparation is complex and lengthy. To this date, data preparation remains the biggest impediment to analytics’ time to market. COGNIMARTS solves this challenge by using its own Artificial Intelligence engine to automate the data preparation cycle. KEY FEATURES ▪Automated discovery and recognition of data entities and relations using ArtificialIntelligence and Machine Learning, even before metadata is provided▪Automated determination and improvement of all possible dimensional modelsincluding scanning of available metadata▪Human-driven evaluation of data models, selection of all relevant ones andcreation of new attributes, business rules, metrics and aggregations▪Automated creation of virtual or materialized Data Marts that can be queried withnatural language (NLP)▪CogniBase – an accelerator to Apache Hive (Hadoop component for SQL queries),nd most of your IT budget on maintaining your current system? Many companies find that constant maintenance eats into their budget for new technology. By outsourcing your IT management to us, you can focus on what you do best–running your business.
Satisfaction Guaranteed
The world of technology can be fast-paced and scary. That’s why our goal is to provide an experience that is tailored to your company’s needs. No matter the budget, we pride ourselves on providing professional customer service. We guarantee you will be satisfied with our work.
NCBR Founding
Projekt współfinansowany przez Unię Europejską ze środków Europejskiego Funduszu Rozwoju
Regionalnego w ramach Programu Operacyjnego Inteligentny Rozwój 2014-2020.
Projekt realizowany w ramach konkursu Narodowego Centrum Badań i Rozwoju: Szybka Ścieżka nr umowy POIR.01.01.01-00-0742/19
Budzet
Wartość projektu: 7 270 261,91 PLN wartość dofinansowania: 5 384 945,97 PLN
TYTUŁ PROJEKTU
Opracowanie i walidacja CogniMarts – systemu opartego na sztucznej inteligencji zdolnego do samodzielnego,automatycznego (bez zaangażowania człowieka) porządkowania i łączenia dużych zbiorów danych
CEL PROJEKTU
Opracowanie i zaimplementowanie nowej technologii CogniMarts, która pozwoli na zautomatyzowanie czynności umożliwiających klientowi analizowanie danych poczynając od przygotowania danych, poprzez modelowanie danych, identyfikację modeli wielowymiarowych oraz tworzenie i zapełnianie wielowymiarowych składnic danych
PLANOWANE EFEKTY
Docelowy system pozwoli na automatyczne przeprowadzanie analiz firmowych baz danych z tysiącami encji (klas obiektów) i budowanie wartościowych informacyjnie składnic danych (marts) wielowymiarowych. Dzięki nim użytkownicy otrzymają szybki i skuteczny wgląd w dane oraz wsparcie decyzyjne, przy czym będzie to możliwe po zdecydowanie niższych niż obecnie kosztach. Możliwe będzie także zasilanie zewnętrznych narzędzi do uczenia maszynowego.
Our Partners
