case study

Automating $18B Financial Transactions per Year

Automation shifts payment processing into cruise control.

KUNGFU.AI developed an AI solution that transformed Triumph by slashing invoice approval time from 178 minutes to an average of 10 seconds while processing $4 billion in transactions and delivering a 10% increase in profit margins.

AI Solution(s)
Change Management
AI Microservices
Industry
Logistics
Cargo truck with cargo trailer driving on a highway

Triumph operates one of the largest transportation payment networks, striving to pay more truckers at top speed.

Triumph logo

Vision

Revolutionize how quickly truckers get paid.

Triumph operates as a factor in the complex transportation finance ecosystem, processing $18 billion in annual financial transactions while managing tens of thousands of invoices daily from small carriers. In this ecosystem, carriers—often mom-and-pop operations with just a handful of trucks—face severe cash flow challenges, waiting up to 90 days for payment after delivering goods for major shippers. Triumph bridges this gap by paying carriers upfront, taking a small fee while assuming the risk of collecting from brokers and shippers.

Beginning in 2023, after recognizing that invoice processing and payment timelines had plateaued, Triumph set out to leverage AI to dramatically improve funding speed while maintaining their invoice risk profile. 

“Funding a client has improved from an average of three hours to a minute or less on loads eligible for Instant Decision. This significantly increases operational efficiency and creates a positive experience for our client.”3Q24 Shareholder Letter

Challenge

Breaking the status quo in a high-stakes, risk-averse industry.

Triumph faced a fundamental tension: diligent review of invoices is critical to mitigating fraud and ensuring payment collection, but speed-to-funding is core to their value proposition. While technology had enabled improvements over the years—from mailing checks to ACH/Wire transfers, reducing funding times from 7-10 days to 24 hours—the rate of improvement had plateaued over the last decade.

However, previous automation attempts using hard-coded systems had failed, typically because stringent rules led to low approval rates. Lacking internal AI expertise, Triumph needed a trusted partner to navigate this challenge. 

Breakthrough

Overcoming past automation failures through transparent "guardrail metrics" and a multi-phase rollout that proved the model could meet or exceed the quality of human-review of invoices.

Working with KUNGFU.AI, Triumph developed and scaled the AI initiative through several carefully orchestrated phases.

Refining the Business Logic: In general, Triumph’s business rules did a good job at capturing risk and preventing negative outcomes. But, we found that not all rules were necessary to perform these functions, and there was a lot of room left to make the ruleset more efficient at preventing bad outcomes. By paring down the ruleset, we were able to simultaneously make the process more efficient while also preserving the institutional knowledge that had served Triumph well for the past several decades. The honoring of existing rules while removing the excess allowed for the remaining rules to better guide decision-making. 

Objective "Guardrail Metrics": To build confidence in the Instant Decision solution, we aligned the organization on three specific negative outcomes to track: Disputed invoices, Short Pays, and Write-offs. By creating transparent dashboards showing these metrics, stakeholders could objectively measure progress as we worked together.

Proving Impact: Through a methodical rollout—from data exploration of $50+ billion in historical transactions to dark launch testing to a 100-day pilot—we demonstrated that the Gradient Boosted Classifier (the model we ultimately chose to implement) didn't just work faster than humans, but actually made equivalent or better decisions because it could analyze all the data points in a way that would be too challenging and time intensive for a human. All carrier invoices currently go through the model, but humans still review edge-cases where the model is unsure.

Outcome

A competitive moat and industry-leading customer experience that exceeded every goal.

The Instant Decision initiative, which completed its rollout to all carrier customers in June 2025, has been a remarkable success across every metric. To date, approximately $4 billion worth of invoices have been processed through the model, with over 50% auto-approved—saving thousands of hours of manual work.

  • Time to payment slashed: Truckers can now be paid for a delivered load within seconds, a speed unheard of in the industry. In one publicized example, a factoring client submitted an invoice on a Saturday morning and was fully funded in 21 seconds—three days faster than normal (3Q24 Shareholder Letter).

  • Material business impact: Triumph’s Instant Decision Model has significantly reduced operational costs, call volumes, and manual review time while improving customer retention. Results include a 57% reduction in short pays, 65% decrease in chargebacks, and 25% fewer disputes in the small carrier segment (1Q25 Shareholder Letter).

  • Lower risk, not higher: By quantifying Triumph’s risk profile and aligning modeling with their rules, the system actually reduces risk exposure.

  • Market differentiation: Triumph is now demonstrably faster than competitors. This is a clear competitive advantage that has been highlighted in shareholder communications and unlocked significant new business opportunities.

“Instant Decision means just what it says. We’ve built a model that ingests data and can make an instant decision without human intervention on whether or not we would purchase this invoice. That’s never been done before… making decisions in less than 30 seconds.”3Q24 Earnings Call

Download the Case Study

quotes icon
“KUNGFU.AI didn’t feel like a vendor—they were true partners. From day one, they brought incredible talent and were people I could jam with, think strategically, and genuinely solve problems alongside. I appreciated their heavy focus on value and ROI—it was never about building tech for tech’s sake. If you’re trying to figure out how to approach AI in your business, I’d absolutely recommend working with them.”
Jason Heilig
Jason Heilig
CTO, Triumph

We sit down with Jason from Triumph to explore how AI is transforming the financial services industry—particularly for clients in the transportation space. Jason shares how Triumph is leveraging AI to make 25,000 real-time decisions every day, improving customer experience while reducing operational costs.

See how KUNGFU.AI’s strategy-led engineering bridges strategy and execution for AI success.

Work with Us

More case studies