Global Fast-Food Chain Slashes Regression Time by 50% Leveraging Coforge’s AI-Driven Test Automation for Automated Order Taker
Overview
The rapid advancement of AI & automation technologies has led global brands to reimagine traditional customer service models. An American multinational fast-food corporation undertook a transformative project to enhance its drive-through experience by implementing an IBM Watson-based Automated Order Taker (AOT) system. The sheer complexity of these requirements demanded a robust quality engineering partner capable of simulating real-world business scenarios and ensuring the reliability and accuracy of the automated system.
Absence of dashboards for real-time QA visibility and point of failure
Need to assess QA processes for AOT & GMA
Solution
Assessed QE processes & practices end-to-end across Agile lifecycle, test automation, and QA governance; mapped team setup, capacity, and utilization; prioritized quick wins for near-term value.
Real-Time Scenario Impersonation: Developed test frameworks that mimic actual car arrivals and departures, automate order-taking processes, and validate order accuracy on the POS system
Technology Integration: Leveraged AWS for vehicle signal detection and Azure Cognitive Services for speech-to-text and text-to-speech conversion, ensuring smooth user interaction and order completion
Dialect Support: Customised test cases to support 16 different dialects, enhancing accessibility for a diverse customer base
Validating orders on the POS kiosk and cross-verifying with menu inputs via real-time ordering API calls
Assessed and implemented automation for AOT, NPOS, and GMA; evaluated AOT integration feasibility (camera, vehicle detection, voice assistants, POS) and flagged process gaps blocking automation
Supporting 4 different environments of the application (Black Widow, WASP, Mantis and Moon Dragon) with 2 Lanes
Implemented persona-based dashboards using Zephyr Squad for real-time QA visibility; baselined a metrics framework and compliance score to drive outcomes and accountability
Enabled content analytics with proactive alerting for faster issue detection
The Impact
~50% regression cycle-time reduction
~90% automation coverage supporting 16 dialects
7,000+ automated test cases
Support across 5 markets on both Android and iOS
Offshore lab set-up with 40+ devices and 2 POS stores
Comprehensive quality to support switching Menu’s (Breakfast, Lunch and Snack)
