Bogdan Beca

A global product information management (PIM) implementation that included

• Integration of a PIM SaaS (Akeneo) solution as a global PIM for multiple eCommerce systems (Magento);
• Event-driven asynchronous integration with Azure Functions;
• A complex routing schema for markets, environment types, channels, locales, and operations scopes (possibility to scale on hundreds of targets);
• Loosely coupled the design with an extensible configuration to allow heterogeneous eCommerce configurations.

This is an enterprise-level OLTP implementation for a pharma ERP. It's 100% transactional, has full data versioning, with auditing (any information can be traced and restored on request), a configurable formula engine, and high scalability.

This is an oil-and-gas production forecast based on unsupervised learning and high-performance similarity matching included are algorithms for validation against overfitting.

It's been proven 15% more efficient than traditional forecasting on historical regressions.

This is a decision support system for work over incremental production. It involved a complex integration among operational systems, a master data store, SAP PM/PS, and SharePoint reports. It also has complex Oracle analytics for multicriteria forecasting, actual and planned time series using dynamic time series shifting, and trend identification.

This state-of-the-art RBI-and-PIM implementation uses a PODS standard model (pipeline open data standard) and ArcGIS for topological processing.

Key Features:
• Linear referencing (with engineering station points and station equations)
• GIS referencing
• Hundreds of parameters for monitoring on each asset for quantitative and qualitative monitoring
• Dynamic segmentation
• FMECA (failure mode, effects, and criticality analysis) risk models for pipelines, flowlines, piping, and static equipment
• SAP PM and MM integration for automation inspection proposal

It has around 100 daily active users (integrity engineers and their teams). We presented the solution at the GITA Oil and Gas Pipeline Conference 2014, Houston, TX, as one of the largest and most complex PODS implementations worldwide.

This is a specialized E&P information management solution that manages several master data marts and legacy data marts. It aggregates around 15 billion records in a daily aggregation window for several hundreds of licensed users.

This was a 10-year-long program where I implemented a massive MDS hub over a standardized E&P data model. This involved integrating a dozen operational systems, SAP PM, and several consumer systems. It exchanges a couple of million pieces of master data daily with extremely strict master data rules; it remains at the core of enterprise operations.

A general-purpose SOM platform that is currently under development. I developed proprietary ML algorithms for SOM unsupervised learning, U-matrix generations, hierarchical clustering (gravitational), and hierarchical segmentation.

I used Azure, SQL, .NET, OpenCL, Vega charts, and .NET MAUI.