Robust duplicate detection
Reworked DynamoDB writes around a transaction-based key with conditional writes, so duplicates are prevented before insert, even on retries.
- Home
- Case Studies
- Whittle Group
Case study · Education Management / Enterprise Systems
Strengthening data integrity for high-volume transaction processing.
Skugal stabilised Whittle Group’s reporting and transaction APIs on DynamoDB and AWS Lambda so every transaction is written exactly once, tested and traceable.
- Client
- Whittle Group
- Industry
- Education Management / Enterprise Systems
- Delivery
- Backend engineering and data-integrity improvements
- Capability
- Cloud & Backend Engineering
The challenge
What Whittle Group was up against.
Whittle’s reporting and transaction services process high-volume data through DynamoDB and Lambda, where accuracy and idempotency are critical. Duplicate writes were inflating reports and forcing manual reconciliation.
- Duplicate transaction writes
- Inconsistent deduplication logic
- Fragile local developer setup
- Thin unit test coverage
- Noisy reporting queries with deprecated fields
The solution
What we built.
Skugal stabilised the transaction pipeline end to end, focusing on correctness, testability and developer experience.
Write-path consistency
Aligned ingestion paths and reduced race conditions in concurrent writes.
Unit tests for transaction APIs
Coverage for duplicate detection, write behaviour and query correctness.
Reporting query clean-up
Removed deprecated fields from transaction queries to simplify payloads and reconciliation.
Local Lambda development
Local execution against a dedicated AWS dev environment, so developers can debug safely without touching production.
How it works
From input to outcome.
- Transaction received
- Idempotent conditional write
- Validated persistence
- Clean reporting queries
- Accurate downstream analytics
Business impact
What changed.
Whittle’s core transaction layer is now reliable: every transaction is correct, testable and traceable, and the team can iterate safely.
Technology
- Node.js
- DynamoDB
- AWS Lambda
- AWS SAM
- Unit testing
- GitHub PR workflow
- Duplicate transaction writes eliminated
- Improved reporting accuracy and consistency
- Higher test coverage and deployment confidence
- Faster developer onboarding and debugging
Good systems don’t just process data. They guarantee correctness under pressure.
Next step
Have a similar challenge?
Tell us about the workflow or product. We’ll come back with an architecture, a delivery plan and where AI agents fit.