AI Purchase Order Fraud Detection Developer

Orvix Works

Remote · $100k – $130k / year · Full-time · Posted 4 months ago

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

This role is no longer accepting applications

Don't miss the next role like this

Create a free profile with your skills. We match you to new AI roles as they open, and companies hiring can contact you directly.

Create free profile, 2 minutes

Similar open roles you can apply to now:

See all open AI jobs

About the role

Procurement fraud — fictitious vendors, inflated invoices, duplicate payments, kick-back arrangements between buyers and suppliers — costs organizations an estimated 5% of annual revenue and goes undetected for an average of 24 months before discovery. The patterns are consistent and detectable: vendors registered to residential addresses, invoices just below approval thresholds, payments to vendors with no prior purchase history, and timing correlations between employee changes and new vendor onboarding. We serve 28 enterprise procurement teams with spend intelligence tooling. Purchase order fraud detection is the risk management capability our clients' internal audit teams are asking us to build. We process $18B in annual purchase order volume — the signal is there. What's missing is the ML system that scores purchase orders and vendor relationships for fraud risk indicators and surfaces the highest-risk transactions for AP team review. You'll build the PO fraud detection system: engineering behavioral and relational features from the procurement transaction network, training fraud risk scoring models on confirmed fraud cases and near-miss signals, and building the detection alert workflow that integrates with procurement approval processes.

Fully remote.

Responsibilities

Engineer behavioral and relational features from vendor registration, invoice patterns, and approval histories
Train fraud risk scoring models on confirmed fraud cases and audit-identified near-miss transactions
Implement vendor relationship network analysis detecting collusion indicators between buyers and suppliers
Build detection alert workflow integrating fraud scores into procurement approval decision flows
Develop fraud pattern library capturing known fraud typologies as detection rule anchors
Monitor false positive rates ensuring fraud flags don't disrupt legitimate procurement operations

Requirements

3+ years in data science or ML with fraud detection, financial audit, or procurement analytics experience
Strong Python, SQL, and scikit-learn skills for structured transaction data modeling
Understanding of procurement fraud typologies, AP controls, and audit investigation methods
Experience building risk scoring systems integrated into operational approval workflows
Ability to work with internal audit teams on fraud case investigation and model validation