A payment is approved for Bob.
Someone changes the recipient to Eve directly in the database.
Would your system stop the payment before the money moves?
Database access should not be payment approval.
That is the idea behind Record Integrity, my independent architecture project.
WHAT THE APPROACH ADDS
1. Check before execution
Compare the payment instruction with an independently stored, accepted original.
2. Find missing records
Detect accepted operations that disappear from the main database.
3. Make retries safe
Prevent repeated processing from creating a second internal debit.
4. Verify the history
Use signed Merkle checkpoints and independently retained evidence to detect changes to recorded audit history.
WHY THIS IS MORE THAN AN HMAC COLUMN
HMAC and Merkle trees are established tools.
The engineering work is connecting them to approval, execution and recovery in an existing system.
My aim is to help companies design and integrate these controls.
The Java/PostgreSQL demo is a reference implementation, not a finished production product. Keys and protected execution must stay outside the source database administrator's control.
READ THE ARTICLE
"Verifiable Record Integrity Without a Blockchain"
Read the attached article, or follow the article link shared with this post.
YOUR PERSPECTIVE
If you work on payments, databases or security:
What would you test first? Where do your existing controls already solve this problem?
---
The views expressed in this article are my own and do not reflect those of my employer.
#DatabaseSecurity #SoftwareArchitecture #ApplicationSecurity