The challenge
What did the client need?
A large insurance company with hundreds of custom applications first hired ioSENTRIX for a penetration test. The issues we found pointed to design and architecture flaws, and to an application security program that was not working.
Engagement details
- Client
- Large insurance company
- Estate
- Hundreds of custom applications, many on-premises and legacy
- Scope
- Testing tools, AppSec processes, SDLC and developer training
- Start
- A penetration test that exposed design and architecture flaws
- Timeline
- Eight months to the headline result, within a three-year roadmap
- Result
- Zero Critical or High open
- Delivered by
- ioSENTRIX application security team
Our approach
How did ioSENTRIX approach it?
More tooling would have added more noise. The turnaround came from tuning what was there, building the processes around it and training developers on the organisation’s own top risks.
The roadmap also added secret, infrastructure-as-code and dependency scanning, application classification and data-handling standards, design review, threat modelling and manual code review, with quarterly business reviews for leadership.
What we found
What did ioSENTRIX find?
Tools that cried wolf, a program that did not exist, and both now fixed.
Theme · ToolingFinding 1
Scanners produced noise, not fixes
The company relied only on static and dynamic application security testing (SAST and DAST). Both were poorly configured, so false positives and legacy debt buried real issues, and developers ignored more than 60,000 reported vulnerabilities.
Business impact. Real risks were lost in the noise, and the tools had lost the trust of the people meant to act on them.
Theme · ProcessFinding 2
There was no program to act on results
There was no working vulnerability management or risk acceptance process. Exceptions in the governance system had fallen out of step with the scanners, leaving a build-up of outdated exceptions.
Business impact. Nobody could say which risks had been accepted, by whom, or which needed fixing first.
The path from scanner noise to zero Critical or High
Program path · from noise to signal
- 1Gap analysisweak points found
- 2Tools tunedfalse positives cut
- 3Programs addedreview, risk, training
Outcome: Zero (Critical or High open in eight months)
The impact
What changed as a result?
Its scanners had produced more than 60,000 open vulnerabilities that developers ignored, about 10% of them false positives. We ran a gap analysis, built a three-year roadmap and started a risk reduction campaign. More than 35% of the Critical findings were closed with platform-level fixes. Within eight months, no Critical or High vulnerabilities remained outstanding.
60K+
open vulnerabilities at the start
0
Critical or High outstanding after eight months
35%+
of Critical findings closed by platform-level fixes
What changed after the program?
What changed across the application estate| Area | Before the program | After eight months |
|---|
| Open vulnerabilities | More than 60,000 reported issues that developers ignored | No Critical or High vulnerabilities open |
|---|
| Tooling | Poorly configured SAST and DAST, buried in false positives | Tuned SAST and DAST, about 10% of findings confirmed false positives; secret, IaC and dependency scanning added |
|---|
| Process | No vulnerability management or risk acceptance; stale exceptions | Vulnerability management, risk acceptance and application security review programs |
|---|
| People | Security seen as noise by developers | Training on the organisation’s top 10 risks and reviews for high-risk apps |
|---|
✓ Verification: Outcome measured across the estate.