The motivation for AIGO

  • Can you trust a computer to take orders for your restaurant?
  • New AI voice ordering solutions
  • Considerations for choosing a phone ordering solution
  • The ChatGPT paradigm shift
  • Introducing Dymos: Dynamic IVR
  • Dymos is database-based
  • Dymos features
  • Conclusion

Introducing Dymos: Dynamic IVR that Takes Phone Orders for Restaurants


Dec 19, 2023

Can you trust a computer to take orders for your restaurant? This is the first question restaurant owners ask themselves when they are considering if they should adopt one of the newly-emerging AI chatbot-style solutions for order-taking. Most restaurants are already using the computer – i.e. cloud services – in some way: to take orders online, through the restaurant chain’s own app or via a 3rd party such as Doordash, Uber Eats or others.

But a new approach has emerged, AI chatbots now exist at some QSR chain’s drive-thru lanes and AI for phone orders is in play. (Of the two, this article focuses on phone orders).

You may want to take a step into the “digital future” by signing up for a service that offers AI to take your phone orders. The reasons? For small stores: you might be having a hard time finding and retaining part-time help to take your phone orders; it has been said that hiring for the order-taking position is one of the most challenging labor activities. Or, your cooks and front-of-house staff may be juggling answering the phone with their other duties. For large chains: you have made a large investment and your call center personnel have been trained on menu and geographic area for delivery; is the call center meeting your expectations to provide quality order instructions? You too might be considering whether AI would provide a better approach or whether it might be an alternative that can supplement the call center.

Here are three key considerations:

  1. Does the AI order service provide consistent quality, i.e. accuracy for order-taking, or will the AI order service somehow mess up a complex food order or delivery instructions so that you end up with unhappy customers (and possibly a bad online review for your restaurant?).

  2. Are there hidden costs? Does the service base it’s price on number of completed orders or on some other approach? The value to your restaurant should be based solely on the number of orders that the service brings in to your establishment. AI is expensive – model training occurs periodically and involves massive costs; inference happens in real-time as the cloud service submits human user turn inputs to AI inference services for various processing tasks. An AI order service may offer an attractive initial pricing scheme but has the vendor provided guarantees that this price level is sustainable?

  3. Does the service provide timely menu information? Technology has entered the “ChatGPT era”, where it’s possible for vendors to leverage the technology of the major platforms (this also includes the generative AI/LLM platforms Google Bard, Microsoft Bing, and others). A vendor may offer to provide a solution that has a strong dependency on a generative AI platform. Although there are ways to coax up-to-date menu and item availability information from the underlying AI platform via sophisticated prompting, keep in mind that these massive AI models are only trained periodically. Will your restaurant’s weekly and even daily updates on things like price changes, specials, hours of operation, etc. be accurately conveyed by the AI to the customer?

When it was first introduced ChatGPT presented a dazzling array of capabilities and indeed has been responsible for a paradigm shift as society now views AI as having at least some seemingly-human powers. Perhaps this would work for restaurants and hospitality as it has for others?

Even prior to ChatGPT many restaurants had become aware of chatbot approaches for order-taking. However recent experience in this area has not produced any clear winner solutions, for drive-thru, for phone, or as voice-enabled chatbot-like interfaces for apps, kiosks or other digital channels.

Introducing Dymos

Software Engineering Concepts, Inc. is pleased to offer a solution to this dilemma. We introduce “Dymos” – “dynamic IVR ordering service”. Dymos is a multi-user cloud service that allows phone customers to choose items by numbers – a “multiple choice” approach for ordering food and drinks.

Dymos can do IVR for food ordering because it has these main breakthrough proprietary advances:

  1. Sophisticated data model for reference data that represents everything about your restaurant that’s needed to handle orders: menu information down to the smallest detail (items, sub-items, modifiers for all items and sub-items), prices, discounts, availability, suggested items, and more.

  2. Proprietary logic that always guides the customer correctly so that they are able to communicate accurate orders that get sent to your restaurant (kitchen and other staff).

  3. Categorical guided ordering experience: when a customer calls in a phone order it cannot be assumed they have your restaurant’s menu handy (the “menu-absent” problem). Thus Dymos always leads off with a list of categories, to allow the customer to drill down to get the items they want. For example a pizza shop Dymos experience might involve “For a pizza press 1, for an appetizer press 2, for a salad press 3, for a drink press 4”.

Dymos is first-and-foremost database-based. This lends it to being “always-up-to-date”. When you subscribe to the Dymos service, your restaurant’s menu information and other “metadata” such as your location, hours, etc. is uploaded once; subsequently this information can be dynamically updated as needed using the online Dymos Dashboard that we provide to you.

Examples of information that you can dynamically update using the Dymos Dashboard:

  • Menu item availability (i.e. to reflect that your restaurant is sold-out of an item or one of it’s ingredients)

  • Price changes

  • Specials and discounts

  • Hours of operation

In addition, Dymos handles pickup and delivery arrangements as needed, and it provides secure payment options to the customer. (You as the restaurant merchant can choose whether or not to enable and/or require customer payment at time-of-order).

Recap: Dymos is IVR for Ordering Food and Drinks

IVR is not a new technology: you are probably familiar with the IVR experience when calling banks, insurance companies, medical office or any of countless other businesses that have used it to their advantage. While it sometimes provides a seemingly-shallow benefit, perhaps only serving to route callers to a human operator, we can all agree on its value and there are usually no doubts as to its accuracy in handling its tasks.

To-date we at Software Engineering Concepts, Inc. are unaware of any competing solution that brings IVR to the restaurant ordering experience. We are excited to be the worlds-first to offer Dymos as this solution.

Availability

Dymos is available as a subscription service; this is ideal for small restaurants and small to mid-sized restaurant chains. For larger chains we offer an API and SDK so that your developers can integrate Dymos into existing infrastructure if desired (contact us to learn how the Dymos API may be leveraged for your situation).

Please contact us at info@softwareengineeringconcepts.com if you think Dymos is a good fit for your restaurant or chain.