Building an in-house security function for a fintech
A specialty-finance company of about 130 people moved security away from an outside IT provider. For six months I led the consulting team that ran it day to day and reported to executives and shareholders. This is how the work was organized and what it changed.
Facundo Montefalcone, lead cybersecurity infrastructure architect. Client name and open findings are withheld.Context
A US specialty-finance company runs on Microsoft 365 with a separate identity provider and a mixed macOS and Windows fleet. It had relied on an outside IT provider for security and wanted the function in-house.
My role was Lead Cybersecurity Infrastructure Architect and Engineer, through a consultancy. I owned planning, prioritization and security reporting to executive leadership and shareholders.
Attribution: I led the team. Engineers on the team executed much of the hands-on work listed here, and I say "we" where that is the case.
First 30 days
- Environment auditInventory of identities, endpoints, applications and admin paths, so priorities came from evidence.
- Enterprise browserArchitecture for a managed browser as a policy enforcement point.
- MDM migrationPlan to consolidate device management onto one platform per operating system.
- SOC 2 readinessMapped controls to evidence and to named owners.
- Privileged accessHardening of admin roles and the paths used to reach them.
What the team delivered
| Workstream | What we did | Result |
|---|---|---|
| Endpoint coverage | Reconciled EDR enrollment against two device-management tools and the asset inventory | 113 devices confirmed, about 14 unaccounted devices traced one by one, status report to leadership |
| Device control | USB mass-storage block piloted with 5 users, executives included, before wider rollout | No exceptions requested during the pilot |
| Leaver hygiene | Audited every deactivated account over the previous quarter | Corrected accounts left suspended instead of deactivated, reassigned directory group ownership, proposed HR-triggered termination notices |
| Risky sign-ins | Investigated flagged identities, revoked sessions, documented closure | About 5 cases closed and the identity security score cleared |
| Data loss prevention | Migrated file policies from the cloud-app security tool to Microsoft Purview and tuned third-party storage rules | About 99% less alert noise. Spotted a licensing change that would make third-party coverage pay-per-event and raised the cost decision months ahead |
| External collaboration | Restricted guest access to an approved domain list. Rerouted operational alerts for 9 external vendor users into a tenant-scoped Teams channel | Those alerts no longer leave the tenant by email |
| Incident readiness | Phishing response runbook: message removal, click investigation, log review | A repeatable first-hour response |
| Network | Firewall firmware upgrade plan built with vendor support and a risk assessment | Presented to leadership before any change window |
| Audit support | Answered SOC 2 auditor requests on network segmentation and change approval | Agreed rule: only changes with business-wide impact need executive sign-off |
How I ran the function
Daily sync, owners and dates
Thirty minutes every morning. Every item had one owner and a date, and blocked items were escalated the same day.
One tracker per initiative
Applications, users, roles and status in a sheet the client could read without a meeting.
Reports for executives
Status written as decisions, risk and cost, with the technical detail attached, not leading.
Written approval first
Anything that could touch the business got written sign-off before the change, and minutes after every meeting.
Related work
The same team then handled the sister organization's exit from its outsourced IT and SOC provider. Read the MSP exit case study.