DiningRD

Revolutionizing Ingredient Import Efficiency with AI

Executive Summary

DiningRD, a leading Dietitian Consulting and Foodservice Software Services provider serving over 4,000 healthcare communities across 46 U.S. states, partnered with ArchitectNow to optimize its backend ingredients-import process. To improve efficiency and free up team resources for product innovation, DiningRD implemented an AI-enabled solution that automated ingredient mapping. This new search-and-match system, integrated seamlessly with existing workflows, reduced manual effort, increased import efficiency, and positioned the company for ongoing scalability with minimal disruption.

Opportunities for Improvement

DiningRD (DRD) manages the ongoing import of approximately 1,500 vendor-specific ingredients each month. Each vendor often uses unique naming conventions, with as many as 30 different variations mapping back to the same master ingredient in DRD’s system. This process required dedicated administrative time for searching and matching ingredients individually. By identifying this as an opportunity, DiningRD set out to streamline the task, reduce repetitive manual work, and better direct team efforts toward advancing product capabilities for long-term care communities.

Business Impact

While the manual approach was effective, it required significant administrative hours and slowed the integration of critical data into the software. As vendor numbers and ingredient volumes continued to expand, DiningRD recognized that automation would create a more scalable process while reducing operational costs. With approximately 250,000 vendor-specific ingredients mapped to about 8,000 master ingredients, efficiency gains were key to supporting both internal teams and supply chain partners nationwide.

Objectives

The goal was to develop an AI-driven solution to automate the nightly import of vendor ingredient back-feeds, reducing the need for manual matching while maintaining oversight. The system needed to integrate seamlessly with existing dietitian-facing software, allow administrators to review matches below a set confidence threshold, and provide scalability for future volumes. Ultimately, DiningRD sought to increase efficiency while freeing its team to focus on enhancing services and introducing new AI-driven innovations.

The Solution

Project Overview

ArchitectNow designed a tailored AI-enabled solution for DiningRD, leveraging Azure's AI Search and LLM-based capabilities. The system automates ingredient matching and processes imports nightly. A human-in-the-loop approach allows administrators to review and approve low-confidence matches, ensuring accuracy. Integrated with DiningRD's React-based frontend and SQL database, the solution preserved workflow continuity while positioning the platform for scalability and ongoing AI-driven enhancements.

Key Features

• Automated Search and Match: AI accurately identifies and maps vendor ingredients to the master list.
• Nightly Import: Processes 1,500 ingredients automatically each night, reducing manual workload.
• Human Oversight: Matches below a configurable confidence threshold are flagged for administrator review. Feedback options (thumbs-up/down) enable the AI to learn and improve.
• Scalable Architecture: Supports growth in vendor and ingredient volumes without sacrificing performance.
• Seamless Integration: Works within the existing React and SQL ecosystem, preserving user experience.

Technology Stack

• Azure AI Search (indexer, synonym search, hybrid search, scoring)
• Azure OpenAI Model (LLM for edge cases)
• Azure Storage Accounts Queue (import job management)
• Azure Functions with Microsoft Semantic Kernel framework (.NET)
• React v18 with Redux Toolkit (frontend)
• SQL Database (secure and efficient data storage)
• Application Insights (monitoring and log analytics)

Lessons Learned

This project highlighted the value of combining AI automation with configurable human oversight. By setting confidence thresholds and enabling ongoing feedback, DiningRD not only streamlined its current operations but also created a framework for continued AI-driven innovation to eliminate manual processes and support future product advancements.