The Problem
Enterprise finance teams lose billions annually to duplicate invoice submissions and vendor fraud. The detection methods in use — manual spot-checks and rule-based systems — were catching less than 30% of fraudulent transactions and couldn’t scale with growing transaction volumes.
Accenture needed a platform that could be deployed across its client base, adapt to different ERP systems, and catch fraud with enough confidence to act on automatically.
The Approach
I joined the platform team during its growth phase and became responsible for client implementations — translating the core detection platform into working solutions for clients across manufacturing, retail, and financial services.
Each implementation required understanding the client’s ERP setup, invoice data structure, and risk tolerance. I built a reusable configuration framework that reduced per-client setup time from 8 weeks to 3.
I also worked closely with the data science team to improve the model’s precision on specific fraud patterns we were seeing in the field — duplicate vendor submissions, split invoicing, and roundtrip transactions.
The Outcome
- $1M+ in duplicate and fraudulent invoices prevented across my direct implementations
- Platform grew from 50 to 200+ enterprise clients during my tenure
- Personally implemented for 40+ clients across North America, Europe, and APAC
- Configuration framework I designed became the standard onboarding approach for the team
What I Learned
Scaling a platform across diverse enterprise clients is fundamentally a communication problem as much as a technical one. Each client had different definitions of “fraud” — aligning on that upfront, before touching any data, was the highest-leverage thing I could do on every engagement.