Building a Real-Time Fraud Detection Platform for Renewable Energy Certificates
To strengthen the integrity and efficiency of Europe’s renewable energy certificate (REC) market, a new high-performance system was developed to handle real-time transaction validation, large-scale data processing, and fraud detection. These capabilities were integrated directly into the existing trading platform, helping to close critical infrastructure gaps and support secure, reliable certificate exchanges across borders.
Meet Our Client
The Association of Issuing Bodies (AIB) is an organization that develops and promotes a standardized system for energy certificates across Europe. It is responsible for the European Energy Certificate System (EECS), which provides a regulated and transparent framework for Guarantees of Origin (GOs) as defined by the EU's Renewable Energy Directive.
The AIB brings together national issuing bodies from dozens of European countries to harmonize their certification systems. Its inter-registry hub connects these national systems, facilitating the secure and reliable transfer of certificates across borders, enabling companies and consumers to reliably track the origin of their energy and make credible claims about their use of renewable electricity.
Case Study Shortcut
Challenge
Transaction latency
REC validation taking over five minutes per transaction created a major bottleneck, frustrating users and creating windows for potential fraud. The goal was to reduce validation time to near-instantaneous.
Lack of scalability
The platform crashed under high transaction volumes, causing operational downtime and halting trading. The system required complete re-architecture for high availability and scalability.
Complex interoperability
The new engine had to integrate into a live, 24/7 cross-border platform connecting multiple national registries, requiring zero-downtime deployment to avoid market disruption.
Goal
The project’s core objective was to develop a fraud detection system that could validate certificate transactions in real time—while preserving strict transfer sequencing rules. This meant integrating a new data and analytics platform with AIB’s operational marketplace application and rebuilding the system architecture from scratch.
- Build a high-performance fraud detection engine
- Ensure strict sequencing and legitimacy of every REC transfer
- Achieve interoperability with existing systems
- Improve transparency, trust, and auditability
- Lay foundation for future AI/ML fraud detection models
- Empower self analysis
Outcome
The upgrade delivered a step-change in performance and oversight. Transaction validation times dropped from over five minutes to just 15 seconds, dramatically increasing operational efficiency and ensuring real-time integrity even during peak trading volumes.
Fraud detection is now fast, reliable, and transparent – capable of flagging suspicious activity in real time with rule-based logic that stands up to legal and regulatory scrutiny.
With centralized access to both real-time and historical data, AIB now has a powerful lens into market behavior, regulatory impact, and emerging vulnerabilities – enabling more proactive, data-driven governance.
The platform also acts as a secure, self-service analytics hub for AIB members. Advanced dashboards and granular access controls empower participants to explore their data with confidence, without compromising security or compliance.
Before
- Transaction validation took over 5 minutes
- Data retained for only one month
- Fraud detection was slow, limited, and hard to scale
- System became unstable under heavy load
After
- Transactions validated in just 15 seconds
- Full transaction history permanently accessible
- Real-time, rule-based detection with clear insights
- Scalable architecture ensures consistent reliability
Integrate those solutions in your company
Case Study Details
Approach
Modernized the architecture
- Rebuilt the legacy backend from scratch using scalable, cloud-native technologies hosted on Microsoft Azure.
Implemented real-time data streaming
- Integrated RabbitMQ for message queuing and Redpanda Connect for high-throughput event streaming to handle continuous certificate transactions across borders
Built a serverless fraud detection engine
- Used Azure Functions to process incoming data in real time, compile complex validation queries, and send them to the analytics engine.
Deployed advanced fraud logic in Databricks
- Leveraged Databricks to implement rule-based, explainable fraud detection that is scalable and auditable.
Enforced strict transaction sequencing
- Ensured every REC transfer follows a valid, time-sensitive ownership chain with no parallel or out-of-order processing.
Enabled long-term data retention
- Archived full transaction history to support audits, compliance checks, and historical analysis.
Added secure, self-service analytics for users
- Built dashboarding and self-analysis capabilities with row- and column-level access control, allowing AIB members to explore their data securely.
Technology
Databricks
Azure
Redpanda Connect
RabbitMQ
Our team
Paweł Żak
Senior Data Engineer
Marcin Marczyk
Delivery Director
Our Team Expert Opinion
We had to engineer a system that combined real-time processing, strict sequencing logic, and large-scale data integration, without disrupting a live trading platform. It was a rare mix of performance, precision, and resilience
Marcin Marczyk Director of Delivery at Addepto
