# Reducing Vulnerabilities on a Lending Platform

> TechSupport Team maintained CredAxis’ Java lending platform ahead of an audit. Over six months, the client’s Snyk records show critical and high CVEs falling from 33 to 5. Recurring defects fell from 24 to 7 per quarter in Jira.

- Publisher: TechSupport Team
- Canonical page: https://techsupport-team.com/case-studies/lending-platform-vulnerability-maintenance/
- Updated: 2026-11-16

Client: CredAxis · Fintech & Payments · UK SME lending platform (95 staff)

## Client and engagement facts

- **Client:** CredAxis
- **Industry:** Fintech & Payments
- **Business context:** UK SME lending platform (95 staff)
- **Delivery model:** Software maintenance
- **Application:** SME lending platform with decision API
- **Supplied geography:** UK
- **Company size in source:** 95 staff

## Application stack and tools

### Application stack

- Java 11→17
- Spring Boot
- PostgreSQL
- AWS

### Tools used

- Jira
- Snyk
- Datadog
- web application firewall

Related services: [Software maintenance](https://techsupport-team.com/services/software-maintenance/)

## Why Support Was Needed

The client needed evidence that linked each change ticket to a pull request, a deployment and a verification check. The work paired dependency updates and code repairs with that record.

## Technical Work

- Linked tickets to reviewed pull requests, deploy records and post-deploy checks.
- Prioritized vulnerability remediation, starting with two publicly known issues.
- Repaired the rounding in loan decisions and added a regression test.

## Representative Incident

An issue in a credit-bureau SDK had no upstream patch. The team added a web application firewall rule for the known path and documented the residual risk in the audit trail.

## Results and how they were measured

Table: Reported results and their measurement scope

| Metric | Baseline | Result | Units | Measurement window | Source and scope |
| --- | --- | --- | --- | --- | --- |
| Recurring defects | 24/q | 7/q | /q | 6 months | Client Jira |
| Critical/high CVEs | 33 | 5 |  | 6 months | Client Snyk |
| Decision API P95 | 1.9 s | 620 ms | Baseline: s; Result: ms | 6 months | Client Datadog |

## Evidence and Limitations

Two vendor SDK CVEs were recorded as accepted risks with compensating controls. The table reports the platform’s measures over six months.

## Vendor Vulnerability Handling

The team recorded compensating controls and documented accepted risks from unpatched vendor dependencies in the audit trail.

## Scope a Similar Engagement

Share the dependencies, recurring defects and release evidence your maintenance work must cover. [Request a support plan](https://techsupport-team.com/contact/) or explore [software maintenance](https://techsupport-team.com/services/software-maintenance/).

## Related pages

- [Software maintenance](https://techsupport-team.com/services/software-maintenance/): The related service for this engagement.

## Contact

Request engineers from TechSupport Team or ask a question: https://techsupport-team.com/contact/
