On a weekday afternoon, an Uphold customer might verify their identity, link a bank account, fund it, buy an asset, and move it to a crypto wallet before the hour is out. Once funds leave on a crypto rail, there is no clawback, so Uphold’s fraud team often has to decide whether that person is who they claim to be before the money is gone for good.
Uphold is a global digital money platform serving customers in more than 140 countries. It lets consumers and businesses hold, exchange, and move both traditional and digital assets. Its wallet supports roughly 300 currencies and connects to payment rails across the US, UK, Europe, and Latin America, from cards and ACH to Apple Pay, PayPal, Venmo, and Pix. Its Topper product gives customers instant on-ramp and off-ramp access between fiat and crypto, and its enterprise API lets established businesses offer crypto to their own customers on Uphold’s regulated infrastructure.
Crypto platforms are held to the standards of regulated financial institutions while moving money at crypto speed.
Uphold is using Oscilar’s Agentic Risk Platform to run fraud and compliance as one operation, with specialized agents now working across both.
TL;DR
The platform: Uphold runs fraud and compliance on Oscilar across roughly 300 currencies and a dozen payment rails.
The challenge: Crypto funds can’t be clawed back once they leave, and Uphold’s fraud and compliance teams were working from separate systems with separate data.
The foundation: Crypto-native models and device and behavioral intelligence helped cut fraud losses 35% year over year, and compliance later joined the same platform.
Agents in action: Agents work the first pass of fraud alerts, draft case narratives, and screen sanctions and PEP alerts using Uphold’s own procedures, with 100% QC on every alert an agent has worked.
What it means for crypto platforms: One record of the customer lets fraud and compliance work as a single decision chain, and agents earn more discretion as their work proves out.
Where Uphold started: Fraud and compliance at crypto speed, on separate systems
Uphold’s risk program has to keep pace with that breadth of currencies and payment rails while meeting the standards of a regulated financial institution. Frank Walton, Uphold’s Global Chief Compliance Officer, argues that crypto’s risk profile is often misunderstood. “People think crypto is a black hole where bad actors operate with impunity. It’s the opposite. The blockchain lets you see things a traditional institution never could, "said Frank Walton, Global Chief Compliance Officer, Uphold
That visibility comes with obligation, since Uphold’s responsibility can extend ten or twenty hops beyond its own walls.
As the business accelerated, the tooling struggled to keep up. Uphold was launching products and entering new regions, and every new product needed fraud controls before it could launch. Rob McCall, Uphold’s VP of Fraud Prevention, was also facing fraud that had changed shape, spread across more actors, devices, and addresses, each used less often. The repeat offenders and clustered IPs a rules engine depends on were exactly the patterns weakening.
The fraud tools McCall’s team tested fell short in one specific way: they would score a customer and refuse to say why. In a business where every wrongly blocked customer can walk to a competitor and every decision has to be defensible to an auditor, that opacity was disqualifying.
Fraud and compliance were also working from separate systems, with separate data and vendors, coordinating through Slack and email. When an investigation crossed from fraud into compliance, the customer picture had to be rebuilt on the other side.
A crypto-native fraud foundation Uphold can explain
Fraud moved to Oscilar first. Oscilar trained Uphold’s models on its own data and customer journey, with Uphold defining the signals and outcomes that mattered, from stolen-payment fraud to first-party fraud and scam victims. Device and behavioral intelligence added another layer, combining IP, device, keystroke, and input signals that had lived in silos.
“What stood out with Oscilar was the explainability, and the customization and agility we could have in developing all of our flows.”
Rob McCall, VP of Fraud Prevention, Uphold
McCall’s team tunes thresholds to each product’s risk appetite without waiting on engineering, which matters when every enterprise API client brings its own customer base. When a new fraud trend began spreading across accounts that looked unrelated on paper, Oscilar’s device and behavioral team connected the activity through harder-to-mask signals and worked alongside Uphold’s analysts to stop it.
The results came quickly. Average handle time for fraud alert review dropped 30% in the first three months, false positives fell 15%, and overall fraud losses declined 35% year over year. More than 150 people at Uphold now work in Oscilar across fraud and compliance, and when a new typology appears, the response can be a workflow change the same day.
RESULTS
35%
YoY reduction in fraud losses.
One platform closes the gap between fraud and compliance
Walton watched the fraud deployment before committing his own program. With those results in hand, he brought compliance onto the same platform, so an alert that starts in fraud can move into compliance with its evidence and history intact in one case.
“An alert can now be handled by both a fraud person and a compliance person, with all the context carrying across.”
Frank Walton, Global Chief Compliance Officer, Uphold
That is FRAML in practice. The same customer, evidence, and history carry from fraud review into compliance action, which keeps that context in one place and avoids handoff errors that used to appear when information was retyped across systems. It also helps Uphold plan. A rise in fraud alerts signals a rise in escalations, and because the platform absorbs the volume, the business can assess a new product’s P&L knowing the compliance cost won’t materially change. Walton sees regulatory expectations moving the same way, toward institutions that use AI and machine learning so their people focus on what matters.
Putting agents to work across fraud and compliance
Because Oscilar built the foundation first, each agent works from Uphold’s customer data, decision history, models, workflows, and cases in one system.
“The agents were the reason we chose Oscilar…We bought the vision, and they’ve delivered on it.”
Rob McCall, VP of Fraud Prevention, Uphold
On the fraud side, an L1 review agent works the first pass of an alert using Uphold’s procedures and broader account context, and a case narrative agent drafts the investigation write-up. Early results show alert resolution time falling roughly 35%.
EARLY RESULTS
~35%
Faster fraud alert resolution with Oscilar agents.
On the compliance side, sanctions and PEP screening agents apply Uphold’s own dispositioning procedures. “It was imperative that the agent matched our procedures, not the other way around," said Frank Walton, Global Chief Compliance Officer, Uphold
Walton set a conservative bar. The agent should close only what was certain to be safe, and missing an OFAC match would have been a dealbreaker. It flagged every true match. Document extraction is already in use in due diligence, and Oscilar’s agents are projected to deliver up to 5x faster screening and EDD document review.
PROJECTED
5X
Faster screening and EDD document review with Oscilar agents.
How Uphold is giving agents room to earn discretion
For McCall, the goal is front-line review that runs automatically so analysts can focus on the true outliers, with people staying accountable through the human oversight built into Oscilar’s agents.
“We’ve done 100% QC on every alert an agent has worked. It’s reliable and it’s provable. The next step is segmenting the alerts and giving the agent more discretion where it has earned it.”
Frank Walton, Global Chief Compliance Officer, Uphold
That discretion starts with repetitive, high-volume work, the kind Walton knows from doing level one reviews himself. The next agents extend the same model. Document translation, expanded due diligence support, KYC, and KYB are coming onto the foundation Uphold already has, instead of arriving as another set of disconnected systems.
“I’m confident we have what it takes to stop whatever fraud comes next.” —Rob McCall, VP of Fraud Prevention, Uphold
See how Oscilar runs fraud and AML as one operation for digital asset platforms. Read the full Uphold case study or speak with an expert.
How Uphold and Oscilar are Building an Agentic FRAML Model for Crypto Risk
Fraud came first, powered by crypto-native models and device and behavioral intelligence. Compliance followed onto the same platform, and now specialized agents work across both sides with Uphold’s procedures, shared context, and human oversight built in.


Oscilar Team
The Oscilar Team is comprised of experts from many domains of risk operations. These articles express viewpoints and knowledge from a variety of sources and contributors across the organization.
DISCLAIMER
The content on this website is provided for informational purposes only and does not constitute legal, tax, financial, investment, or other professional advice. Any views or opinions expressed by quoted individuals, contributors, or third parties are solely their own and do not necessarily reflect the views of our organization.
Nothing herein should be construed as an endorsement, recommendation, or approval of any particular strategy, product, service, or viewpoint. Readers should consult their own qualified advisors before making any financial or investment decisions.
Oscilar makes no representations or warranties as to the accuracy, completeness, or timeliness of the information provided and disclaims any liability for any loss or damage arising from reliance on this content. This website may contain links to third-party websites, which Oscilar does not control or endorse.








