LCP
Overview

A BERT-based NLU model helped to create Voicebot for food deliveries and curbside pickups, along with the voice-based navigator for the nearest pickup.

At A Glance

industry
Industry
Hospitality
region
Region
USA
duration
Duration
8 Weeks

Technical Stack

python
nlp
node
audiocodes
spacy
dialogflow-cx

Client Profile

The client is an American table service diner-style restaurant chain established nearly 80 years ago with over 1000+ locations around the world serving breakfast, lunch, and dinner round the clock.

Challenge

  • As many of its restaurants are located on highways and freeway exits, it is challenging for customers to place an online order through the website or mobile app while driving for curbside pickup.
  • Restaurants are open 24x7 and a large number of orders float in round the clock, so there was a need for agents to be available 24x7 to reply instantly for better customer service.
Architecture diagram showcasing NLU-based voicebot for food deliveries, curbside pickups, and navigation

Solution

  • Designed Voicebot using Machine Learning algorithm which understands user voice input and using NLP, it's converted it to text which the computer can understand to place the order
  • Easy access to nearby restaurant location and get a step by step voice-activated direction to reach faster at the location
  • Place the order over the phone in advance to save time for curbside pick up
  • Integrated menu items with a third-party system to keep menu items in sync for all the platforms
  • Integrated the payment gateway for customer ease to complete order at one go over call
  • Instant customer support round the clock for any queries related to customer orders
  • Transfer the issue to the agent at any point in time in case the bot is not able to understand the customer request
  • Upsell and cross-sell food items based on personalization and past order history
  • Train and deploy model continuously based on user input received

Key Benefits

  • 20% average growth in orders within 6-months of system launch
  • Reduction of service agent's average response time by 60%

Develop your next idea with usGet in touch