F88 — From a cautious pilot to a governed hybrid-cloud operating model

F88 runs a high-scale lending business across more than 800 branches in Vietnam. Its customers depend on reliable digital and branch operations every day, while its infrastructure team manages a hybrid estate across AWS and on-premises systems.

As the business grew, daily infrastructure work became harder to scale. The team had to manage resource inventory, cloud cost, health checks, compliance controls, incident response, and operational standards across many environments.

F88 did not begin with a large transformation program. It began with a practical question every regulated finance team asks before trusting automation with production:

Can this be verified?

CloudThinker earned trust the same way reliable infrastructure does — one small, controlled, auditable step at a time.

F88 was not going to hand production access to an AI agent based on a promise.

The team needed automation, but it also needed control. Any system operating across a regulated hybrid estate had to be scoped, auditable, and safe by default. It had to work across AWS and on-premises infrastructure. It had to produce evidence, not just recommendations.

So CloudThinker started with the lowest-risk, highest-friction work the infrastructure team handled every day.

The first pilot was read-only. CloudThinker built a live inventory of resources, ran resource checks, identified drift and waste, and performed routine health checks across critical services.

Nothing changed automatically. Every result came with a clear explanation: what was checked, what was found, why it mattered, and what action the team should consider next.

That was the point.

F88 could compare CloudThinker's findings against what the infrastructure team already knew to be true. The results held up. The team trusted the checks, then trusted the workflow, then expanded the scope.

Once the pilot proved reliable, F88 expanded CloudThinker from a read-only assessment tool into a governed operating layer for its hybrid infrastructure.

CloudThinker connected to F88's cloud and on-premises environments through scoped, brokered access. Every action ran through identity, approval, and audit controls. The platform did not replace the infrastructure team's judgment. It gave the team a safer way to scale that judgment across the estate.

From inventory to standards

The first step was visibility.

CloudThinker mapped F88's hybrid footprint end to end: resources, accounts, environments, services, dependencies, and operational health. This gave the team a shared view of what was actually running, not just what was documented.

From there, F88 encoded its infrastructure standards into reusable CloudThinker Skills: naming conventions, tagging policies, resource guardrails, health checks, cost controls, and operational runbooks.

Instead of relying on manual follow-up, tribal knowledge, or one-off scripts, the same standards could now run consistently across every account and environment.

From daily chores to automation

With standards in place, F88 began automating repetitive daily operations.

CloudThinker now handles much of the routine work that previously consumed the infrastructure team: resource checks, drift detection, cost recommendations, service health checks, compliance evidence collection, and common remediation workflows.

Roughly 80% of daily manual operations now run through CloudThinker, with human approval required where risk, policy, or production impact demands it.

The infrastructure team moved from checking dashboards and chasing tickets to designing better architecture, improving capacity planning, and tuning the policies that govern automation.

Faster incident response with Deep Response Engine

When production issues occur, CloudThinker Deep Response Engine correlates signals across F88's observability stack and investigates likely root causes in minutes.

Instead of manually jumping between dashboards, alerts, logs, and infrastructure context, the team gets a structured incident summary: what changed, what broke, what systems are affected, and what remediation path is recommended.

Depending on F88's Auto Mode policy, CloudThinker can notify the team, open a change for approval, or perform an approved remediation under predefined guardrails.

Every action is written back into the audit trail, giving F88 a clear record of what happened and why.

With CloudThinker, F88 now runs its hybrid infrastructure through a single governed operating layer.

Daily operations are automated. Incidents are investigated faster. AWS spend is continuously optimized. Compliance checks run on a schedule, with evidence written directly into the audit trail.

The cost impact was immediate: continuous optimization helped reduce AWS spend by 30% without sacrificing performance. Resource checks and automated recommendations helped the team identify waste earlier and act before cost drift became a monthly surprise.

The operational impact was just as important. Critical lending applications are monitored 24/7 across cloud and on-premises systems. Incidents that once required days of manual investigation can now be root-caused in minutes. The infrastructure team spends less time firefighting and more time on architecture, resilience, and capacity planning.

Compliance also changed shape. Instead of preparing evidence during quarterly audit cycles, F88 runs mitigation runbooks continuously. Controls are checked regularly. Findings are tracked. Evidence is written back into the audit log. The team can prove its posture on any given day, not only during audit season.

The deepest sign of trust came after adoption: F88 helped improve the product.

Running CloudThinker in a regulated, hybrid, multi-account finance environment surfaced requirements most teams never encounter. F88's feedback helped strengthen CloudThinker's hybrid coverage, security controls, approval workflows, and day-to-day infrastructure automation.

Those improvements now benefit every CloudThinker customer.