In this episode of the Restaurant AI Podcast, Matt sits down with Justin Foster, Co-Founder and CRO of Incept AI, for a deep and technical dive into the future of Voice AI in restaurants, particularly in the drive-thru environment.
With years of experience in both restaurants and AI, Justin brings rare insight into what it really takes to build voice technology that works in the chaotic, noisy world of restaurant operations. Matt presses Justin on the tough questions—what’s real, what’s hype, and why so many voice AI companies are entering the space now.
Key Topics Covered:
The evolution of voice AI from “tree-and-branch” rule-based systems to large language models (LLMs)
Why the McDonald’s Apprente experiment failed—and what’s changed since
The four layers of voice AI: raw audio, transcription, intent processing, and TTS
The critical role of clean audio input and why denoising is the bottleneck
How hardware, digital base stations, and cloud-based audio ML interact
Why latency and naturalness in conversations are essential to guest experience
The hidden cost of relying on humans-in-the-loop to "fix" broken AI orders
Why scaling Voice AI requires retraining, not just installing software
How economics, accuracy, and guest trust define the winners in this space
Predictions about moving from cloud to on-prem edge computing in the next 3–5 years
The future of voice AI in training, loyalty, upselling, and personalized service
Visit Incept AI Online: https://www.incept.ai/
Connect with Justin on LinkedIn: https://www.linkedin.com/in/justinfoster/
Visit ClearCOGS Online: https://www.clearcogs.com/
Connect with Matt on LinkedIn: https://www.linkedin.com/in/matthewjwampler/
01:48 Voice AI, explained
03:37 Early innovations in Voice AI
7:45 Challenges of Voice AI
13:05 Latency in Voice AI
15:10 Walkthrough on Voice AI in a drive-thru
19:58 The critical role of clean audio input
24:36 Minimizing mistakes and maximizing accuracy
28:50 Foundation model advancements
30:46 Moving from Cloud to On-Prem
32:08 Model training and differentiation
37:30 Why purchasers are wary of Voice AI
46:57 Robotic vs human interractions
49:14 Which voices do humans prefer?
51:08 The economics of Voice AI
58:17 Do customers recognize Voice AI?
1:01:17 The future of Voice AI
1:06:57 Impact on minimum wage employment
1:07:54 The restaurant industry is resistant to change
1:08:49 Weirdest moments of Voice AI
1:09:47 How it all gets stitched together
1:13:58 Authenticity and regional dialect
14:55 Accuracy in Voice AI and non-Voice AI
1:18:33 Justin's passion for the technology
1:21:57 Contact Justin and Outro