Six months after delivering our first integrated version of the framework, so-called, Minimum Viable Product (MVP), the I-BiDaaS consortium released the 1st Prototype of the I-BiDaaS solution. One of the project main aims is to create new opportunities for self-service analytics towards a comprehensive paradigm tailored to big data analytics. The new version takes this into consideration including many important new features that we have been developed during the second year of the project (2019). The MVP that shipped at M12 (2018) has been built for early demonstration and dissemination of the capabilities of I-BiDaaS. The MVP covered mainly one use-case provided by CAIXA about finding relations between bank customers for security checks, given a set of their IP address connections to online banking services.
You can find listed below the main improvements archived for the new version:
- Enhancement of Batch analysis processing through better algorithms, better data and resource management, and better testing.
- Enhancement of Stream analysis processing through analyzing real-time data via complex event processing.
- Enhanced visualization through I-BiDaaS centralized dashboard where different stakeholders can operate with 3 different platform modes.
- Synthetic Data Fabrication tool provide artificially generated data that has the same structure and statistical properties than the real one, without exposing sensible information.
- Docker support at deployment time.
Through the I-BiDaaS dashboard, which is the common entry point for different stakeholders that are willing to interact with our solution, currently, we are providing 3 different modes to operate the platform, grouping different end-user profiles depending on their knowledge or business needs:
- Expert mode allows developers to upload their own source code providing a flexible solution taking full advantage of the capabilities of the big data technology enablers offered by our solution.
- Self-Service mode allows us to define an analytic pipeline using a set of predefined algorithms that a non-IT expert user can select to perform their analysis in a guided way.
- Co-Developed mode is more business-oriented; the end-users receive support and guidance from I-BiDaaS members in order to customize the analytics pipeline and enhance the visualization of the experiment results.
The 1st I-BiDaaS prototype release has been a major milestone in our roadmap. It has been demonstrated successfully in the scope of the 1st Project Review, held in Luxemburg, for pilots using both synthetic and real tokenized/anonymized datasets.
The following is a list of the experiments achieved within this period:
- Analysis of relationships through IP addresses (Batch processing) [CAIXA]
- Analysis of relationships through IP addresses (Stream processing) [CAIXA]
- The production process of aluminum Die-casting (Batch processing) [CRF]
- Quality of Service in Call Centers (Stream processing) [TID]
The list of experiments is going to be extended during the third year of the project when the 2nd I-BiDaaS prototype has been planned (M30).
If you need further details about the 1st I-BiDaaS prototype, contact us directly through our website’s contact page.