.png)
Sam Nadler (00:00):
Okay, so pretty rhythmic, kind of a cool idea. And then, as the store gets busier, you can hear and see the occupancy, and as you keep going, the party gets higher and higher. And at the top, it’s even got a closing mode. So that’s how we close it out, and then I’ll bring the music back down. Chill.
Sam Nadler (00:23):
Hey everyone, and welcome to Built This Week, the podcast where we share what we're building, how we're building it, and what it means for the world of AI and startups. I'm Sam Nadler, co-founder here at Ryz Labs, and each week I'm joined by my friend, business partner, and co-host, Jordan Metzner. And this week we have a special guest joining us: Angie, the CEO of Standard AI.
Angie Westbrook (00:58):
Thank you, thank you — happy to be here. In one sentence: I’m Angie, CEO of Standard AI, and we use next-generation, cutting-edge computer vision to bring online-level metrics into brick-and-mortar retail. I love building, we have a great team, and with all the attention on generative AI right now, it’s exciting to highlight what computer vision can bring to physical retail.
Sam Nadler (01:27):
Amazing. This week, like always, we’re going to cover a tool we built in the last seven days — but this time, we built it with Standard AI in mind. Maybe it’s even a product feature you could adopt. Then we’ll get into what Standard AI actually does, and finish with the latest AI news: Gamma’s massive fundraise, SoftBank pivoting from Nvidia to OpenAI, a bit about 11 Labs, and one of our favorite tools. Jordan, any thoughts before we jump in?
Jordan Metzner (02:10):
Super excited to chat with Angie — another big week in AI. Never a dull moment, lots of announcements from top-tier companies. Let’s get into it.
Jordan Metzner (02:55):
So the idea here actually started with you, Sam. I just took your thought and put pen to paper, thinking specifically about Standard AI. I’ve been loving AI music generation — I’ve tried Audio, 11 Labs APIs, and others. It democratizes music production: you describe a vibe, and the model creates a whole track. And because Standard AI focuses on retail, we thought: how do we combine music + computer vision + AI to build something a retail store could actually use?
Jordan Metzner (03:38):
If you’ve ever used Spotify’s AI DJ, it takes your playlists and remixes them. Cool sometimes, annoying other times. But we thought — retail stores could go way beyond that. Most stores just play radio hits. The music doesn’t change whether it’s busy, empty, morning, night, raining, or sunny. So we built an AI DJ that changes the music based on the store’s occupancy.
Jordan Metzner (04:50):
We generated five songs — 90 BPM up to 130 BPM and back down — all mapped to how many shoppers are in the store, from “opening mode” to “packed” to “closing mode.”
Jordan Metzner (05:42):
Let me jump into the demo.
Jordan Metzner (06:04):
Here’s the intro track — it even has a welcome message for shoppers.
(plays music)
Jordan Metzner (06:38):
And as occupancy increases, the BPM rises… until you hit the peak, a full-store track.
Jordan Metzner (06:53):
Then we have closing mode — bringing the energy way down.
Sam Nadler (07:09):
Angie, curious for your first thoughts on our retail-mood DJ.
Angie Westbrook (07:28):
I love it — I’m a sucker for music, so easy sell! But Sam, something you said is interesting: you have a hypothesis about how music might influence shopping behavior. That’s exactly where this would pair with Standard AI. Customer experience is everything, and lots of factors affect how people shop. If we played different music, we could actually measure how behavior changes in real time — not just rely on sales data, which lags. I’d love to try this in a store.
Jordan Metzner (08:32):
Exactly. And we even thought about iterating on which tracks work best. Maybe disco performs great in the afternoon, but one disco track does better than the other two. You could keep improving until you find the perfect “yield-maximizing” track.
Angie Westbrook (08:51):
Totally. You could personalize by aisle or time of day. And honestly — creating unique, on-brand music for a store is huge. Every mall plays the same holiday playlist. That shows there’s no data behind it.
Sam Nadler (09:14):
So… no more Michael Bublé on repeat?
Angie Westbrook (09:36):
Don’t take away Bublé — I need at least one song!
Sam Nadler (09:58):
Okay, perfect segue. Angie, earlier you described Standard AI as “Google Analytics for physical stores,” which instantly made the concept click. Walk us through it.
Angie Westbrook (10:32):
Online unlocked rapid A/B testing, personalization, impressions, click-through — all thanks to rich behavioral data. Physical retail? Historically a black box. You only see sales data, which lags and is influenced by season, weather, traffic, and randomness. Now, with computer vision on standard security cameras, you can finally get the same level of behavioral fidelity inside stores.
Sam Nadler (11:01):
And it delivers insights faster than sales logs.
Angie Westbrook (11:54):
Exactly. Sales are the ultimate lagging indicator. Retailers often need 12–16 weeks to run a proper A/B test across 40 stores. With Standard AI, some customers get statistically significant results in three weeks — sometimes using only a couple stores. They can immediately see whether shoppers react the way they expect, instead of waiting months.
Sam Nadler (14:45):
How does it work on a technical level?
Angie Westbrook (15:11):
We’re privacy-first. No facial recognition. We label 26 body keypoints, turning each shopper into a digital stick figure. That allows us to track movement, gaze direction, interest, engagement — without identifying the person.
Angie Westbrook (16:43):
One of our biggest unlocks is the Visual Engagement Score — think of it like an online impression, but for physical shelves. Say you put a new product out. If it doesn’t sell, you don’t know whether people saw it and didn’t care, or whether they simply never noticed it. We give you that middle-funnel visibility.
Angie Westbrook (18:47):
And because the system runs continuously, we can also model predictions, simulate changes before making them, and personalize at scale. It’s really early days, but the possibilities are massive.
Angie Westbrook (20:17):
Also, fun fact — 85–90% of retail sales still happen in-store. So investing in physical retail matters a lot.
Sam Nadler (21:12):
Let’s shift to news. First: Gamma raises a huge $68M Series B at a $2.1B valuation. Their AI presentation tool is booming — 70M users, 30M decks a month.
Jordan Metzner (22:21):
Wild numbers. You’d expect Microsoft or Google to dominate slides forever — but a small startup innovated faster. Shows how incumbents don’t always move at startup speed.
Angie Westbrook (24:23):
Exactly. They reimagined the workflow, not just the file format. And I love that they’re profitable — lean teams doing real innovation.
Sam Nadler (25:44):
Next: SoftBank sells its entire Nvidia stake again and pivots to OpenAI. Big move toward the application layer.
Jordan Metzner (26:38):
Classic SoftBank — always betting on the next wave. And OpenAI truly feels like a hyperscaler now.
Sam Nadler (27:38):
Last one: 11 Labs launches the Iconic Voice Marketplace — brands can license AI voices from Michael Caine, Maya Angelou, Babe Ruth, Mark Twain, Amelia Earhart, and others.
Jordan Metzner (28:36):
It’s the early days of talent monetizing their voice via AI. Even estates of historic figures can participate. And alive celebrities can consent and earn from it.
Angie Westbrook (29:42):
It also raises big questions about consent, rights, and verification. But it’s absolutely where we’re headed.
Jordan Metzner (30:09):
Imagine: Fatman Scoop or Diplo as your in-store DJ — powered by AI. Endless possibilities.
Sam Nadler (30:25):
Amazing episode. Thank you Angie for joining us. Everyone, like and subscribe — and check out Standard AI online.
Angie Westbrook (30:47):
Thank you both — so fun.
Sam Nadler (31:05):
Next week on Built This Week: Axiom Math joins us. See you then.
