Redesigning fraud operations for trust at scale.

Synctera provides the infrastructure that enables banks and fintechs to launch and operate regulated financial products. Within this ecosystem, fraud operations teams review alerts, investigate high-risk activity, and make decisions that directly impact customers, partner banks, and regulatory posture.

As the platform scaled, fraud operations became increasingly complex. Analysts managed growing alert volume across KYC, transaction monitoring, and compliance workflows while coordinating with multiple external partners. Decisions were time-sensitive, auditable, and difficult to reverse.

The challenge wasn't simply moving faster. It was maintaining confidence that decisions reflected the current reality of an investigation — especially when work was blocked, incomplete, or dependent on others.

Despite multiple tools and alerts, analysts lacked a reliable way to understand case state, priority, and ownership — leading to premature closure, blocked work, and invisible risk.

Case state didn't reflect reality. Cases had no clear sense of where they were in the investigation lifecycle. In several workflows, the only way to move a case forward was to mark it complete even when work was still ongoing — so in-progress investigations could appear resolved, while stalled work became invisible.

Ownership was unclear. There was no reliable signal showing who was actively working a case or whether progress was blocked. Multiple analysts could unknowingly work the same investigation, duplicating effort and creating conflicting decisions.

Progress moved outside the system. Blocked cases were resolved through Slack messages, emails, or phone calls. If documentation arrived by email and wasn't uploaded, cases stalled silently — none of it reflected back into the system of record.

The system must always reflect the current truth of work, even when progress is blocked.

To address the pain points, I reframed fraud case management around a single principle — the system must always reflect the current truth of work — and rebuilt the experience around three moves: consolidating fragmented tooling, reducing coordination noise, and rebuilding case visibility at scale.

Creating a single source of truth without replacing core tools.

Fraud investigations used to force analysts to jump between products like Hawk AI, Onfido, and Dotfile, piecing together identity checks, risk signals, and decisions across multiple tools. I mapped how those systems were actually being used and redesigned the workflow into a single internal experience that kept the investigation anchored in one place.

Instead of replacing existing tools, I created a simple in-context way for analysts to open third-party systems when needed, then return to the case to capture findings and add notes without losing progress. This reduced back-and-forth, removed manual reconciliation, and gave teams a clearer, more reliable view of what was happening in each case.

Analysts needed timely updates without constant interruption. Rather than broadcasting notifications, I designed alerts to activate only when intent was explicit — such as mentions or active watching.

For teams that relied heavily on Slack, case notifications were connected directly to shared channels, anchoring coordination back to the system of record while preserving focus.

The dashboard made priority, ownership, and workload immediately legible — without requiring analysts to infer urgency from tables or tribal knowledge.

Before the redesign, urgency was inferred socially. Cases older than a few days were assumed to be urgent, and workload lived in analysts' heads rather than the system.

The redesigned dashboard encoded priority directly into the interface. Cases were ranked using priority signals and manager assignment, letting analysts focus on the highest-risk work without manual sorting.

At a glance, it answered three questions — what needs attention now, who is working on what, and where work is getting stuck — replacing tribal knowledge with shared situational awareness.

Impact was validated by tracking how cases moved through the system before and after launch — focusing on time spent in active investigation and the reduction in stalled or prematurely closed cases. Partnering closely with Operations and conducting post-launch analyst interviews confirmed that improvements reflected real workflow change, even as overall case volume increased.

The dashboard functioned as a workload stabilizer for fraud operations — not a reporting surface.

If another designer took over this system, I'd caution against casual changes to the dashboard. It wasn't just a place to view tasks; it became the primary mechanism through which analysts understood their workload, progress, and capacity.

By making assignment, priority, and active work visible at both the individual and team level, it helped analysts plan their day, set expectations, and avoid the constant feeling of falling behind. This visibility didn't just improve throughput — it reduced burnout by replacing uncertainty with clarity.

At a team level, shared visibility let work be redistributed before pressure became unsustainable. The dashboard shaped behavior, morale, and trust — so any future changes should be validated against real operational behavior and evaluated for their impact on workload balance and team health.