From a legacy SQL/Power BI tool to a serverless, AWS-driven forecasting and arbitrage intelligence engine.
1. Starting Point: A Fragile Legacy Stack
The organisation originally relied on:
- A single SQL database living on ageing on-prem hardware.
- Power BI dashboards stitched together manually.
- Slow refresh cycles, limited modelling, and no ability to forecast or simulate market exposure.
Energy traders, procurement teams, and executives were forced to make multimillion-dollar decisions using rigid tools that couldn’t handle:
- Spot price volatility
- Network loss-factor behaviour
- Contract exposure
- Probability scenarios for risk and arbitrage
The system was functional but fundamentally capped. It required a full rebuild if it was going to support modern energy procurement.
2. Transformation: Rebuilding the Entire Stack in the Cloud
The project rebuilt the legacy system into a fully cloud-native, serverless architecture powered by AWS.
Key achievements:
Serverless data layer
- Migrated all SQL workloads into AWS-managed services.
- Implemented event-driven ingestion for NEM data, network tariffs, contract updates, and market prices.
- Created multi-year storage for market history, enabling long-term modelling without performance degradation.
3. Advanced forecasting & probabilistic modelling
The platform can now simulate:
- Forward spot price curves
- Network behaviour under different seasonal and demand conditions
- Contract vs spot exposure
- Probability-weighted arbitrage opportunities for traders
This moved the business from descriptive analytics to predictive and probabilistic intelligence — something the old stack could never do.
4. Enterprise Impact: A Procurement Intelligence Engine
The new platform became the central procurement and reporting system for multiple major clients across banking, corporates, and nonprofit organisations.
Key outcomes delivered:
1. Procurement Strategy Modelling
- Multi-year scenario simulations
- Contract vs spot risk comparison
- Hedging and arbitrage opportunities projected with confidence intervals
- Exposure reporting for executives and boards
2. Network Behaviour Analytics
- Real-time insights into network loss factors (DLF/MLF)
- Regional and site-level risk scores
- Predictive alerts for material changes impacting future bills
3. Intelligent Invoicing & Validation
- Automated ingestion and cross-validation of invoices
- Detection of mismatches, anomalies, and billing errors
- Serverless reporting distributed automatically to stakeholders
4. Enterprise-Grade Reporting
- Custom BI dashboards refreshed in real time
- Compliance reporting for auditors and regulators
- Automated summaries for procurement committees
The platform transitioned from “a reporting tool” to the decision engine for energy procurement.
5. Commercialisation & Enterprise Service Layer
What began as an internal tool evolved into a full enterprise intelligence service offering:
- Data infrastructure as a managed service
- Forecasting and modelling engine as a subscription
- Invoicing validation and anomaly detection
- Procurement dashboards and strategy reporting
The service brought forecasting, invoice validation and reporting into one platform, replacing the fragmented legacy workflow.
6. Why This Build Stands Out
Built from a tiny start The project started with one Power BI file and an old SQL server. No team. No budget. No roadmap.
Evolved into a modern, elastic platform Cloud-native, event-driven, modular, and capable of ingesting and analysing years of energy market behaviour.
Direct commercial value
- Enabled better procurement decisions across multiple million-dollar portfolios.
- Reduced billing errors and prevented cost leakage.
- Created a repeatable energy intelligence product that enterprises purchased as a service.
Future-proof architecture Serverless, scalable, and designed for forecasting, sensitivity analysis, and market simulation — something few enterprise energy tools can handle even today.
7. Summary
This project represents the classic arc of a well-executed digital transformation:
Legacy → Cloud → Intelligence → Enterprise Product
- Started as a rigid SQL + Power BI reporting stack.
- Rebuilt into a cloud-native forecasting and procurement intelligence engine.
- Commercialised into a recurring enterprise service adopted by large organisations.
In the end, it became the invisible brain powering procurement decisions, invoicing accuracy, risk management and long-range strategy — a quiet but critical engine in the business operations of major Australian enterprises.