Recommendation Engine - Provider Matching

Recommend the best service providers based on an office's location, proximity, contract recency, quality ratings, availability, and other business dimensions. Data lives across ERP and business intelligence platforms.

A facilities management organization needed to streamline its process for selecting service providers for office locations. The decision involved multiple factors, from geographic proximity to contract history to quality scores, and the data supporting those decisions lived in separate systems that did not talk to each other.

Challenge

The core difficulty was not the AI model. It was defining what "best provider" means in different contexts. An office in one region might prioritize proximity, while another might prioritize contract recency or quality ratings. The business logic governing provider selection was multi-dimensional and context-dependent, and the data needed to support those decisions was spread across Tableau and SAP.

What We Scoped

We designed a Python-based recommendation service with a FastAPI server layer, integrating with Tableau for visualization and SAP for provider and contract data. The AI component handles multi-dimensional scoring across location, quality, availability, and relationship history. The estimated AI effort is 120 hours, focused on encoding business rules into a scoring system that reflects how the organization actually makes provider decisions.

The Opportunity

Provider selection that currently depends on institutional knowledge and manual lookups can become systematic and consistent. The challenge is encoding what "best provider" actually means in different business contexts, and that encoding work is where the value of the engagement lives.