Teaching a Computer to Speak | Voice AI in the Real World episode artwork

EPISODE · Aug 5, 2026 · 20 MIN

Teaching a Computer to Speak | Voice AI in the Real World

from ServiceNow Insights · host ServiceNow Community

Voice AI sounds simple until you try to deploy it at scale — across airports, accents, languages, and thousands of employees at once. In this episode of ServiceNow Insights, host Bobby Brill sits down with Midam Kim, an ML engineer and linguist at ServiceNow, to unpack what it actually takes to build voice AI as an enterprise product. From the out-of-vocabulary problem (why AI still struggles with names) to why turn-taking in conversation is a linguistic skill most people never think about, Midam breaks down the human science behind the technology.  In this episode: - Why voice AI is replacing typing as the default way to interact with enterprise systems  - The difference between building for employees (B2B) vs. building for their customers (B2B2C)  - Why an airport is one of the hardest possible environments for voice AI — and what ServiceNow does about it  - The "out-of-vocabulary" problem: why AI still struggles with names, accents, and rare expressions  - Why ServiceNow's secret sauce is hiring linguists, not just engineers  - The linguistic framework behind every voice interaction: sounds, words, and turn-taking  - Why voice AI is like teaching a kid to speak for the first time Chapters00:00 — Welcome to ServiceNow Insights00:22 — Meet Midam Kim, ML Engineer & Linguist00:35 — Why voice is replacing typing01:48 — What voice AI actually does for employees03:40 — B2B vs. B2B2C: who's really using this?05:11 — Desk employee vs. airport traveler: two different problems06:42 — Building for an ever-changing environment08:53 — Why airports are the hardest use case09:54 — Accents, fluency, and the diversity problem11:20 — "My Name Is. My Name Is. My Name Is." — the OOV problem12:50 — The coffee shop name story13:20 — How ServiceNow trains its models14:59 — The 3 linguistic layers: sounds, words, interaction16:38 — Midam's turn-taking story from Korea18:33 — Why voice agents can't be "that person you avoid"20:16 — "We can make it great"  Subscribe for more ServiceNow Insights episodes on AI, voice technology, and enterprise innovation. Related episode: Voice AI Agent Evaluation — how ServiceNow measures whether voice AI meets human expectations. https://youtu.be/x7Ks932T18o For more about voice in AI from Midam Kim - https://youtu.be/3NUf6W_FMWs?is=wGc7BfyhiDp8JlOW  #VoiceAI #EnterpriseAI #ServiceNow #ArtificialIntelligence #Linguistics #ConversationalAI #AIProduct #PodcastSee omnystudio.com/listener for privacy information.

Episode metadata supplied by the publisher feed · Published Aug 5, 2026

Embed this episode

NOW PLAYING

Teaching a Computer to Speak | Voice AI in the Real World

0:00 20:50

No transcript for this episode yet

We transcribe on demand. Request one and we'll notify you when it's ready — usually under 10 minutes.

No similar episodes found.

No similar podcasts found.

Frequently Asked Questions

How long is this episode of ServiceNow Insights?

This episode is 20 minutes long.

When was this ServiceNow Insights episode published?

This episode was published on August 5, 2026.

Can I download this ServiceNow Insights episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.
URL copied to clipboard!