Redesigning fraud operations for trust at scale.

Synctera is the infrastructure banks and fintechs use to launch and run regulated financial products. Inside that world, fraud teams review alerts, investigate risky activity, and make calls that affect customers, partner banks, and the company's standing with regulators.

As the company grew, fraud work got harder. Analysts were handling more alerts across KYC, transaction monitoring, and compliance while coordinating with several outside partners. Every decision was time-sensitive, auditable, and hard to walk back.

The goal wasn't just speed. It was trusting that a case reflected what was actually true right now, especially when the work was blocked, half-finished, or waiting on someone else.

With plenty of tools and alerts, analysts still couldn't reliably tell a case's status, priority, or owner. The result was premature closures, stuck work, and risk no one could see.

A case's status didn't match reality. There was no clear sense of where a case sat in its lifecycle. In some workflows the only way to advance a case was to mark it complete, even when work was still going. So active investigations looked resolved, and stalled ones disappeared.

Ownership was murky. Nothing reliably showed who was working a case or whether it was blocked. Two analysts could pick up the same investigation without realizing it, duplicating work and reaching conflicting decisions.

A lot of the real work happened off-platform. Blocked cases got sorted out over Slack, email, or a phone call. If a document came in by email and never got uploaded, the case quietly stalled, and none of it made it back into the record.

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

I anchored the redesign to one principle: the record should always reflect the current truth of the work. From there I focused on three moves: pulling the scattered tools together, cutting the coordination noise, and making case visibility hold up as volume grew.

Creating a single source of truth without replacing core tools.

Investigations used to make analysts bounce between Hawk AI, Onfido, and Dotfile, stitching together identity checks, risk signals, and decisions as they went. I mapped how those tools were actually used, then rebuilt the workflow into one internal experience that kept the whole investigation in a single place.

Rather than replace those tools, I gave analysts a simple way to open them in context, then come right back to the case to record findings and notes without losing their place. That cut the back-and-forth, removed manual reconciliation, and gave everyone a more reliable picture of each case.

Analysts needed to stay informed without being interrupted constantly. So instead of broadcasting every notification, alerts only fired on a clear signal, like a mention or a case someone was actively watching.

For teams that lived in Slack, case notifications flowed into shared channels, pulling coordination back to the record without pulling people out of their work.

The dashboard made priority, ownership, and workload obvious at a glance, so analysts didn't have to guess urgency from a table or from memory.

Before the redesign, urgency was a social guess. Anything more than a few days old was assumed urgent, and people carried their workload in their heads instead of the tool.

The new dashboard put priority right into the interface. Cases were ranked by risk signals and manager assignment, so analysts could go straight to the highest-risk work instead of sorting it themselves.

At a glance it answered three questions: what needs attention now, who's working on what, and where things are stuck. That replaced tribal knowledge with a picture everyone shared.

I measured impact by watching how cases moved before and after launch, mainly the time spent in active investigation and the drop in stalled or prematurely closed cases. Working closely with Operations and interviewing analysts afterward confirmed the gains were real changes in how people worked, even as case volume climbed.

The dashboard was a workload stabilizer for the fraud team, not a reporting screen.

If another designer picked this up, I'd tell them to be careful with the dashboard. It wasn't only a task list. It was how analysts understood their workload, their progress, and how much they could take on.

By showing assignment, priority, and active work at both the personal and team level, it helped analysts plan their day, set expectations, and shake the feeling of always being behind. It didn't just move more cases. It eased burnout by trading uncertainty for a clear picture.

Across the team, that shared view let work move around before pressure boiled over. The dashboard shaped how people behaved, how they felt, and how much they trusted the tool. Any future change should be tested against real operational behavior and judged by what it does to workload balance and team health.