Customer Challenges

The organization faced a growing data quality crisis as its global operations generated massive volumes of low-quality accounts payable transactions—spread across multiple currencies, languages, and frequently missing essential attributes.

Their data classification was largely semi-manual, slow, and error-prone, relying on a cumbersome rule-based engine with over 15,000 rules, yet achieving less than 70% accuracy. This created huge blind spots in spend analysis, making it impossible to view the bigger picture under one roof. The current process was costly, inaccurate, and not effective.

Partner Solution

1. AI-Driven Classification Engine

Ibexlabs deployed an AI-powered classification workflow leveraging Amazon Bedrock and Amazon Nova Models to automatically classify accounts payable transactions with 95% accuracy. The system handles multilingual and multi-currency inputs and replaces the customer’s previous rule-heavy process with a fully automated model-driven approach.

2. Human-in-the-Loop (HITL) Review Interface

To maintain accuracy and allow continuous improvement, the solution includes a Human-in-the-Loop (HITL) interface where users can review and refine classifications. HITL feedback supports exception handling and enables more reliable outcomes for transactions requiring human oversight.

AWS services used: Amazon S3, Amazon Bedrock Knowledge Base, Amazon SES, Amazon SNS

3. Real-Time Processing with Enterprise Rule Integration

The platform supports real-time processing and integrates with the customer’s existing enterprise rules. This ensures that automated classifications are aligned with internal controls and that flagged items can be reviewed quickly.

AWS services used: Amazon EKS for processing, AWS ELB for load balancing, and Amazon CloudWatch for monitoring.

4. Scalable ERP Integration Across 200+ Categories

The solution integrates with the customer’s ERP systems and supports classification across 200+ spend category taxonomies, ensuring consistent categorization and streamlined downstream processing.

Supporting AWS services:
AWS WAF, Amazon Route 53, AWS Identity and Access Management (IAM), AWS Secrets Manager

Lower Total Cost of Ownership

The AI-driven classification engine significantly reduced the total cost of ownership for procurement spend analysis by eliminating manual effort, reducing rework, and streamlining category assignments across global operations.

Higher Classification Accuracy

With 95% accuracy, the solution cut misclassifications to less than 5%, ensuring cleaner data, more reliable reporting, and improved downstream decision-making for procurement and finance teams.

Increased Productivity for Category Managers

The automated workflow saved over 150 person-hours for category managers, reducing time spent on manual validation and allowing teams to focus on higher-value procurement activities.

Advanced AI Insights for Procurement Leaders

A built-in AI insights tool now equips procurement leaders with sophisticated “what-if” analysis, enabling more informed planning and faster decision cycles across sourcing, supplier strategy, and spend optimization.

Better Supplier Consolidation & Risk Management

With accurate, consistent classification feeding directly into corporate finance systems, leaders now have clearer visibility into supplier landscapes. This enables more effective supplier consolidation, improved risk management, and tighter alignment with enterprise-wide financial reporting.