Portfolio: Senior Technical Architect

1. Project Identity

WEIS (Well Engineering Information System) is a mission-critical enterprise platform designed for Shell’s Well Engineering division. It serves as the primary backbone for managing the lifecycle of drilling and completion activities, business planning (Busplan), and performance improvement initiatives (PIP). The system acts as a strategic “Decision Support System” that bridges high-fidelity engineering data with financial forecasting, enabling global stakeholders to track operational efficiency and project benchmarks in real-time.

2. Architectural Challenges

• Highly Variable Data Domain: Unlike standard CRUD applications, well engineering data is deeply nested and non-uniform. Each well has unique phases, maturity levels, and technical risk indices, making traditional relational schemas rigid and fragile.
• Legacy Data Harmonization: The system required bi-directional synchronization with legacy Oracle-based enterprise systems and SAP ERP modules. Maintaining data integrity while mapping disparate data structures was a significant hurdle.
• Computational Intensity: Generating complex engineering visualizations (e.g., Waterfall Charts, Probabilistic Time/Cost Estimations) requires processing massive datasets through multi-step algorithms (Learning Curve Factors, NPT, and TECOP performance modeling).
• Operational Continuity: The platform needed to support long-running background tasks (Report generation, Batch uploads) without impacting the responsiveness of the real-time engineering dashboards.

3. Decision Logic (Trade-off Analysis)

• Persistence: Why MongoDB over SQL Server?
  • Analysis: Well engineering phases are dynamic; new attributes (e.g., specific risk indicators or environment-specific data) are added frequently.
  • Decision: Chose MongoDB to leverage its schema-less nature. This allowed for rapid iteration of the domain model without the overhead of complex SQL migrations or polymorphic table structures.
  • Trade-off: Accepted the lack of multi-document ACID transactions (at the time) in favor of high write throughput and data structure flexibility.
• Frontend: ASP.NET MVC + Knockout.js
  • Analysis: The user base required a desktop-like experience for data-heavy grids and charts.
  • Decision: Implemented a hybrid approach using Knockout.js for client-side MVVM bindings and Kendo UI for specialized widgets. This provided a responsive UI while maintaining the SEO and security benefits of server-side MVC.
• Communication: SignalR for Real-time Feedback
  • Analysis: Long-running calculations for “Business Plans” often exceeded HTTP timeout limits.
  • Decision: Integrated SignalR to provide a persistent websocket connection, allowing the server to push progress updates and calculation results to the UI asynchronously.

4. Business Impact

• Reduced Planning Cycle: Automated the generation of “Monthly Late Estimates” (MLE), reducing the time required for engineering leads to update project statuses from days to hours.
• Improved Forecast Accuracy: The implementation of automated Data Quality Checks (QC) and cleansing algorithms significantly reduced manual entry errors, leading to more reliable financial forecasting.
• Centralized Engineering Intelligence: Replaced thousands of disconnected spreadsheets with a single “Source of Truth,” enabling global benchmarking across different operating units (OUs).
• Operational Scalability: The system successfully scaled to manage hundreds of wells globally, supporting Shell’s expansion into more complex deep-water drilling environments.