Team Ignite Insights · Oct 20, 2026 · 13 min read

Ignite Startups: How VoiceBit Is Transforming Restaurant Ordering with AI with Jay Patel | Ep303

VoiceBit co-founder Jay Patel shares how AI is transforming restaurant ordering, recovering lost revenue, and helping restaurants reclaim customers from delivery apps.

Restaurants are losing orders, revenue, and customer relationships to unanswered phone calls and third-party delivery platforms. VoiceBit wants to change that with AI.

For years, restaurants have relied on platforms like DoorDash and Uber Eats to handle digital orders and deliveries. While these services provide convenience, they also introduce significant commission costs and create distance between restaurants and their loyal customers.

Jay Patel, co-founder and CTO of VoiceBit, believes artificial intelligence can help restaurants regain control of their ordering channels without forcing them to overhaul their existing operations.

In a conversation with Brian Bell, founder of Team Ignite Ventures and host of the Ignite Podcast, Jay explained how VoiceBit evolved from a voice AI experiment into a broader restaurant ordering platform.

Their discussion revealed several lessons about building AI products, identifying overlooked business problems, and creating technology that customers actually want to use.

From Microsoft Azure AI to Restaurant Automation

Before co-founding VoiceBit, Jay Patel developed his technical expertise through software engineering roles at Visa and Microsoft.

At Visa, he worked on loyalty and benefits systems that analyzed large volumes of transaction data to deliver targeted customer offers. The experience introduced him to distributed computing and processing data at massive scale.

He later joined Microsoft, working on Azure support AI systems designed to automate responses to common technical issues.

The challenge was straightforward: Microsoft received enormous volumes of support requests, many involving problems with known solutions. Rather than requiring engineers to handle every request, Jay's team worked on using generative AI to identify existing solutions and provide them directly to customers.

That experience helped shape an idea that would eventually become VoiceBit.

If AI could automate repetitive support interactions for a global technology company, why couldn't similar technology help small businesses answer customer calls?

Restaurants became the natural starting point.

Unlike industries requiring specialized professional knowledge, restaurants were accessible. Jay and his co-founders could walk into local establishments, speak directly with owners, and test their technology in real operating environments.

VoiceBit subsequently emerged from Berkeley SkyDeck and closed an oversubscribed pre-seed funding round led by 1984 VC.

The Failed Pizza Order That Proved AI Voice Ordering Could Work

One of VoiceBit's earliest experiments produced a surprising result.

The team deployed an early version of its AI ordering agent at a Berkeley pizza restaurant.

A customer called and successfully interacted with the AI to place an order.

But there was a problem.

The customer had ordered a half-and-half pizza, and the system failed to capture the toppings correctly. The restaurant received an incomplete order, forcing the team to call the customer back for clarification.

From a technical perspective, the experiment had failed.

From a customer adoption perspective, however, it revealed something important.

The customer believed the order had been placed successfully and reported liking the experience.

That changed Jay's assumptions about how people would interact with AI.

He initially worried that customers might reject the idea of talking to artificial intelligence when ordering food. The experiment suggested that customers cared more about whether a service was available and convenient than whether a human or machine answered the phone.

The first lesson was not that AI ordering had become technically perfect. It was that customers were willing to use it even when the underlying technology still needed improvement.

For an early-stage startup, that distinction mattered.

Why Missed Restaurant Calls Represent Invisible Lost Revenue

Restaurants face a particularly difficult operational problem during peak hours.

Friday evenings, for example, often bring a rush of customers ordering food while kitchen employees are already handling preparation, payments, and in-person service.

When multiple customers call simultaneously, staff may place them on hold or miss their calls entirely.

Some customers call again. Others simply order somewhere else.

The challenge is that restaurant owners often cannot measure the revenue associated with orders they never received.

Jay described this as one of VoiceBit's biggest early sales obstacles.

Restaurant owners were skeptical because they did not know how many orders they were losing.

Instead of relying entirely on sales pitches, VoiceBit began running short pilots lasting one or two weeks.

The results allowed restaurants to observe how many calls the AI handled, including simultaneous orders during busy periods.

For owners, seeing multiple orders processed in parallel made the problem tangible.

VoiceBit's system can answer multiple calls simultaneously, communicate in different languages, take orders, collect payments, and arrange delivery dispatch.

It also integrates with restaurant point-of-sale systems, including Toast, Clover, and Square.

Orders appear in the restaurant's existing operational workflow, reducing the need for employees to learn an entirely new system.

The objective is to automate ordering without disrupting the kitchen.

The Bigger Problem: Restaurants Are Losing Control of Their Customers

VoiceBit initially focused on voice automation, but Jay eventually recognized a larger opportunity.

During the COVID-19 pandemic, third-party ordering and delivery platforms became essential for many restaurants.

These services provided the logistics and digital infrastructure that businesses needed to keep serving customers.

However, this convenience came at a cost.

Jay explained that restaurants can face platform commissions of approximately 20% to 30%, creating pressure on margins.

More significantly, customers who previously ordered directly from a restaurant increasingly began ordering through third-party marketplaces.

That shift changes the restaurant's relationship with its customers.

A customer opening a delivery app to order pizza may also encounter advertisements and recommendations for competing restaurants.

Even loyal customers become accessible to competitors within the same marketplace.

Jay believes restaurants should prioritize protecting their recurring customers and providing direct ordering experiences that preserve those relationships.

This insight encouraged VoiceBit to expand beyond phone calls into web ordering.

Instead of simply replacing a human phone operator, the company began building infrastructure to help restaurants manage their own ordering channels.

VoiceBit offers usage-based pricing and a fixed monthly subscription option, with no commissions on orders according to Jay.

The larger ambition is to help restaurants recover direct customer relationships while reducing their dependence on intermediaries.

How AI-Powered Upselling Can Increase Restaurant Revenue

VoiceBit's value is not limited to answering calls that employees might otherwise miss.

Jay reported that restaurants using the platform can see approximately 5% to 8% in upsell revenue lift.

The mechanism is relatively simple.

When customers order food, especially during busy periods, they may forget complementary items such as drinks, dips, or sides.

Human employees may also lack the time or training to recommend relevant additions consistently.

An AI agent can access menu details, available inventory, customer preferences, and previous orders.

It can then make contextual recommendations during the ordering process.

For example, when someone orders a pizza, the system may suggest an appropriate drink or side based on what is available and what the customer typically purchases.

Unlike a newly hired employee, the AI does not need to memorize the menu or remember individual customer histories manually.

Jay sees this ability to combine operational knowledge with personalized recommendations as a significant source of value for restaurant owners.

The broader implication extends beyond restaurants: an automated interaction becomes more useful when it can complete a transaction and improve the customer's experience, rather than merely answer questions.

The Surprising Metric That Became Misleading

As VoiceBit expanded, Jay began monitoring several indicators of product adoption.

One metric mattered particularly to him: AI usage.

The reasoning was that declining usage could indicate customers were abandoning automated interactions or requesting human assistance.

But VoiceBit encountered an unexpected problem.

AI usage fell even though order volume did not decline.

The explanation was not customer dissatisfaction.

It was a product improvement.

The team had introduced a memory layer that allowed its AI agents to recognize returning customers and remember their previous orders.

Instead of explaining an entire order again, a regular customer could simply confirm their usual items.

Jay described how an order could be completed through a handful of affirmative responses.

The conversations became dramatically shorter.

Consequently, total AI usage declined even while the system continued processing orders.

This revealed an important measurement problem.

A metric that previously indicated customer engagement could become misleading when the product grew more efficient.

For founders building AI products, the lesson is significant: more AI usage does not necessarily mean more customer value.

Sometimes the strongest product improvement makes the underlying technology less visible.

Why VoiceBit Expanded Beyond Voice Agents

As competition in restaurant voice AI increased, Jay began reconsidering what would create a defensible product.

His conclusion was that voice technology alone was insufficient.

Building an impressive conversational agent might be technically exciting, but restaurant owners need much more than a system capable of understanding speech.

They need completed orders, payment processing, kitchen integrations, accurate menu information, and reliable delivery arrangements.

These components are less glamorous than conversational AI, but they determine whether the product works in a restaurant.

Jay acknowledged that VoiceBit initially made the mistake of concentrating too heavily on the voice bot itself.

The team subsequently invested in the operational infrastructure necessary to support an entire ordering transaction.

This distinction also shapes Jay's view of competition.

He does not believe restaurant voice AI must become a winner-take-all market.

Different restaurants have different needs, and multiple companies can compete by offering different combinations of capabilities.

VoiceBit's strategy is to remain close to restaurant owners, observe how they operate, and improve the complete ordering experience.

Rather than competing solely on conversational sophistication, the company aims to make its technology operationally indispensable.

Why Phone Ordering Is Far From Dead

Before entering the restaurant industry, Jay assumed phone ordering was becoming a relatively minor channel.

Digital ordering websites and delivery apps appeared to be replacing traditional calls.

VoiceBit's experience challenged that assumption.

Jay said phone ordering and web ordering generate approximately equal order volumes across the company's deployments.

That finding was particularly surprising because VoiceBit introduced web ordering after developing its phone automation product.

According to Jay, the two channels together now account for more than half of the revenue at restaurants where VoiceBit is deployed.

The lesson is that apparently outdated customer behaviors may remain economically important even when technology trends suggest otherwise.

A business does not necessarily need to eliminate traditional interactions.

Sometimes it needs to make them more convenient.

ChatGPT Ordering and the Future of Restaurant Commerce

VoiceBit is expanding into additional ordering channels, including ChatGPT.

Jay does not claim that customers will inevitably abandon conventional ordering methods in favor of AI assistants.

Instead, he sees enough early indications of shopping activity through ChatGPT to justify preparing restaurants for that possibility.

His goal is for VoiceBit to become an interaction layer connecting restaurants with customers across multiple direct-ordering channels.

That includes phone calls, websites, and potentially AI-powered interfaces that customers increasingly use to discover products and services.

VoiceBit's websites are also designed to be interactive rather than static.

Customers can communicate with the website and ask it to perform ordering-related actions instead of navigating conventional menus.

Jay believes the long-term direction of restaurant technology will involve greater automation, stronger service availability, and pressure on the commissions charged by intermediaries.

If restaurants can manage customer relationships directly through increasingly capable AI infrastructure, third-party platforms may face greater competition.

That possibility, however, remains Jay's outlook rather than an established industry outcome.

Why Startup Founders Must Unlearn Big Tech Habits

Moving from Microsoft to VoiceBit required Jay to change how he approached product development.

At a company like Microsoft, engineering teams can depend on established infrastructure and extensive internal systems.

Startup founders cannot make the same assumptions.

They must build the tools, processes, and operational foundations that larger companies already possess.

But the most important change involved the development process itself.

At Microsoft, Jay had been accustomed to ensuring systems worked accurately end to end before releasing them.

At VoiceBit, he learned to reverse that sequence.

Instead of waiting for a polished product, the team could deploy a proof of concept, observe customer reactions, and work backward from what happened.

The early Berkeley pizza experiment illustrated why this approach mattered.

A technically imperfect solution revealed enough customer demand to justify continued development.

Had the team insisted on perfect accuracy before testing the product, it might have delayed discovering whether customers wanted the service at all.

Jay also emphasized the importance of remaining willing to change one's assumptions when evidence contradicts them.

His beliefs about AI adoption, phone ordering, and restaurant technology all shifted through direct customer interactions.

The Overlooked Benefit of Multilingual AI

Beyond operational efficiency and revenue, Jay sees another underappreciated opportunity in conversational AI.

Restaurants serve customers from different linguistic and cultural backgrounds.

Hiring employees who can understand every customer's preferred language or accent is difficult, particularly for smaller businesses.

AI voice agents can help address this limitation.

Jay argued that multilingual capabilities allow restaurants to make their services accessible to a broader range of customers without requiring separate phone operators for different languages.

This advantage is not simply about lowering labor costs.

It expands who can conveniently interact with a business.

For Jay, that accessibility is among the most promising outcomes of applying AI to customer-facing services.

The Bigger Lesson: Solve the Business Problem, Not the AI Problem

Jay Patel's journey from enterprise AI engineering to restaurant automation illustrates a distinction that many technology startups overlook.

A sophisticated AI model does not automatically translate into a valuable business product.

Customers rarely purchase technology simply because it is technically impressive.

Restaurant owners want calls answered, orders processed, payments collected, and customers served without unnecessary friction.

VoiceBit's evolution reflects that reality.

The company started with an AI voice agent but expanded into payments, POS integrations, web ordering, personalized customer interactions, and additional commerce channels.

Its early experiences also challenged several conventional assumptions: customers were willing to order from AI, phone ordering remained substantial, and declining AI usage could indicate a better product rather than a failing one.

For founders and investors evaluating emerging AI companies, these lessons point toward a fundamental question.

Does the technology merely demonstrate what AI can do, or does it solve a complete problem that customers already care about?

Jay's experience suggests the most enduring AI businesses may be those that make the underlying intelligence almost irrelevant to the customer.

The technology succeeds when the order simply gets placed.

Chapters:

00:01Introducing Jay Patel and VoiceBit
00:49From self-taught programmer to Microsoft AI engineer
05:40The idea behind VoiceBit
08:33The first AI pizza order and its unexpected lesson
09:46How VoiceBit automates restaurant phone orders
12:10Reclaiming customers from third-party delivery apps
14:28Why restaurants must own their customer relationships
16:00VoiceBit's pricing and commission-free model
16:58Measuring AI adoption and customer engagement
19:14How conversational memory changes ordering behavior
20:41Convincing restaurants through real-world pilots
22:34AI-powered upselling and revenue growth
24:21Turning restaurant websites into interactive AI experiences
25:22Why phone ordering remains a major revenue channel
26:19Competing in the restaurant voice AI market
29:28Building the infrastructure for direct restaurant ordering
31:03Why VoiceBit launched ChatGPT ordering
32:20The future of restaurant margins and AI productivity
34:59Moving from Microsoft to startup ownership
36:21Why startups should ship before achieving perfection
37:24What changed Jay's mind about customer trust in AI
38:31Why restaurant owners don't want AI for its own sake
39:00Staying curious and being willing to be wrong
41:03Multilingual AI and restaurant accessibility
43:27Why computer science fundamentals still matter
44:34Jay's interest in building an AI-powered hedge fund
46:00Where to find Jay and VoiceBit
46:40AI existential risk and Jay's 10% P(doom) estimate
50:39P versus NP and the future of computation

Listen to this episode

Open the full episode page ↗

Read the full transcript

Brian Bell (00:01.373) Hey everyone, welcome back to the Ignite podcast. Today I am delighted to have Jay Patel on the mic. He's the co-founder and CTO of VoiceBit, in fair disclosure, a portfolio company. It's an AI ordering platform that helps restaurants win back the phone, web, and delivery orders that's been bleeding them to third party apps. before he started VoiceBit, he spent about a year leading Azure support AI work at Microsoft, where his team used generative AI to deflect thousands of support cases. so there's definitely some transfer learning there to use a machine learning term. voice bit voice bit came out of Berkeley Skydeck, Go Bears, and just closed an oversubscribed pre seed round read by led by nineteen eighty four VC. Thanks for coming on, Jay.

Jay (00:41.309) Hey, thanks for having me.

Brian Bell (00:42.865) Yeah, good to see you again. well, I usually kick it off with the question what is your origin story? What's your background?

Jay (00:49.816) Yeah, sure. so my name is Shay, originally from India. I self-taught myself programming when I was in school. And it was pretty interesting because back in those days computer was computers were used to call us a brain. And it was always intriguing to me like what exactly goes on inside those boxes, right? And so learning how to code actually gave me that appreciation on how these things work. And because of that, I just did my bachelor's in computer science. I learned a lot of things about this. Sorry, mark that. learned a lot about all of this theory of computations, the operating systems, the information theory, all those advanced advanced topics. So after that, I worked probably a three years as a software engineer back in India. and that's when I learned a lot about practical software engineering. So like how the things work in production environment, how the things is built, maintained, deployed in production environment. And that's when I learned that's when I actually realized how it actually separates from the real world. and and what I learned back in the theory, right? after that I came to United States to do my masters in computer science. and I think one of the most interesting subject that I learned here in Castle LA was I think two of them. One was the information theory and another one was the reinforcement learning. And I'm talking about 2019 because it was way before all of this AI boom started. And the the moment I learned about this Markov decision process, MDT MDP, I'm not sure if you know about this, but it's it's this the center point about MTP is that It's just whatever state you are in right now, that's what matters. Nothing in the past matters. And so the current state holds the entire history. and that's all you have to know about in order to predict the next state, the next best at. And so that was the that was the aha moment for me. Like I I think it totally makes sense. why didn't I not even think about it before? (03:14.858) but after that I joined Visa to do my again resume my professional computer science or the software engineering career. there I was working on the loyalty and benefits platform. So you see all these offers in your Cheese. I'm not sure if you use Cheese, but you see the offers on your Cheers app. Those are the ones being targeted to you by the loyalty platform. And so the way it works is that it analyzes your spending habits and then then segments you. And then that's how they target you. And so it was an amazing experience there, like how to process petabytes of data and trying to figure out how the things work in terms of distributed computing. After that, I moved to Microsoft to work on Azure. and it was a completely different landscape there. like like the the services that we were maintaining were like a global scale, like billions of requests every every day and the idea there was I mean the platform that I was working on mainly was dealing with the Azure support AI and what that means is let's say if you if you have any issues with your sub subscription on the Azure let's say you're not able to see the data in your in your in your storage account. Usually those require some specific set of steps. They are mostly predefined. Like you don't really need to be a rocket science to know that. However, when you are doing it for the first time you definitely need to know about those things. But when you think about the Azure scale, it's like millions of requests like this. And you just cannot afford to have an engineer working on those requests. And so the idea was that you need to figure out based on the request from the user, if we already know such solutions from our knowledge base, then how can we even give those solutions to the user rather than them creating a ticket and then having an engineer work on it and then back and forth. So that was the thing that I worked there. but yeah, and after that I (05:35.16) Started the voice book.

Brian Bell (05:37.073) How did you come up with the idea for VoiceBit?

Jay (05:40.301) Yeah, so I think it was around twenty twenty five when this whole AI boom was anyways on its I would not say the peak, but at least in a decent space. I was seeing I mean I I think at during during that time Entropy came up with the sonnet models and it was around the same time when I was also exper experiencing personally the the the benefit of coding agents and and and and in general the one of the okay I I would just rephrase it. mark that I would say.

Brian Bell (06:20.817) Yeah. Yeah, go for it. Yeah.

Jay (06:29.388) I think around that time during twenty ninety five, I would say If you if the goal of AI is to is to help boost the productivity. That is the biggest the that is the biggest win. Exactly. I see that as an L leverage. Now there are a bunch of ways you can do do that. One is you can just for example the way Anthropic does is just doing the coding work. what I thought about is the a a lot of this

Brian Bell (06:46.769) Yeah. It's a it's a tool that provides leverage. Yeah.

Jay (07:06.572) businesses that you are not able to interact with through phones just because of lack of that labor, right? And the idea was that because we were using Alexa devices a lot, we thought why can't we just interact with any business over Alexa? Why can't we just do that using this technologies? I should be able to just call any business and and talk to them, like have the services. And that was the moment. Exactly. Exactly. Exactly. And so that was the that was the thing, and that was the core center center idea about the whole voice pit. And the the the restaurants were the most obvious one, and in fact the most accessible ones for us because we don't really have any other background like the like the doctors or anything like that. So

Brian Bell (07:35.037) Yeah. What time how late are you guys open? Yeah, like can I make a reservation? All that stuff. Yeah.

Jay (08:03.552) Those were the ones where we knew that we could actually walk in and pitch the idea. And so that's how this whole wise bit idea wise bit idea came along. I think we we pitched this to

Brian Bell (08:16.537) the aha moment, right? Because you're you're basically doing the same thing at Microsoft, but Microsoft doesn't run a pizza shop, obviously. You know, Microsoft runs a a massive corporation with lots of support tickets. What was that like aha moment where you're kind of like, I could actually apply this to what like restaurant owners, what they're dealing with?

Jay (08:33.888) Yeah, so I think before even we thought about implementing the voice bed to the restaurants, my personal biggest fear was that would people even talk to the Airbots to place a pizza order? And when we actually deployed this like poorly crafted solution to one of the Berkeley shops, what happened is it actually I mean, we were able to talk to the customer. I mean the customer was able to place an order. Through the agent. However, it was a half and half order and there was no toppings on it on the receipt. And so obviously, I mean we just messed up there. And we basically had to call the customer to get the actual order. However, the issue the the thing that convinced us was that the customer actually thought that they were able to place an order, that they were successful in placing an order and they actually liked it. We actually asked the customer afterwards. And that was the moment we thought, you know what, I think this should work. This could work. Because for them it's the it's it's a matter of availability of the service and not not necessarily who they are talking to.

Brian Bell (09:46.77) What's the integration process look like? for somebody who's never heard of VoiceBit, maybe there's a restaurant owner listening to this podcast right now and they're thinking, Well, I'd love to have AI answer my phones and and take a pizza order. How does it work? And what's that process like?

Jay (09:54.445) Yeah. (10:02.604) Yeah, so I think imagine a imagine a restaurant maybe during Friday, during peak hours. And during peak hours th those are interesting hours because those are the hours where you get most of the orders and you're

Brian Bell (10:17.285) Yeah, you call you call the pizza place, you're like, Hey, I want to make a order. Or can you hold, please? And then like you're like fifth in line, so you're on hold for five minutes while they take the other four pizza orders before they can get to you, right?

Jay (10:21.654) Yeah, exactly. (10:27.899) Exactly, exactly. And so that's the a usual scene in any restaurant. and and and a lot of people might relate if they're watching. And so what ViceBit does is that it can it basically makes her phone silent. And so all the orders, all the all the calls that that people are making to a restaurant to place an order or make an inquiry, for example, they all route it to the VoiceBit agent. it can take any, any, any number of calls in parallel, in any language. It can just take an order end to end, collect the payment, dispatch a delivery if it's a delivery order. And so all you get in the kitchen is just a usual ticket in your KTAS, in your POS, like a like any other order. And so that's how the voice bit works.

Brian Bell (11:19.613) And then so do do you have to integrate with the restaurant systems or I'm I'm I'm guessing if you're integrating with mom and pop pizza shops, for instance, they don't like sometimes they're just kind of writing it down and giving it to the cook and saying, Yeah, large, you know, pepperoni and sausage sausage pizza, go.

Jay (11:38.233) Yeah, yeah. So exactly. I mean, we do integrate with a lot of rest a lot of POS companies. So I mean you know all of them, like the most of them, Toast, Clover, Square, those are the major ones. And so we we integrate with most of them, and that's how we are able to integrate with their existing system and they don't have to change their workflows at all. Like the orders from the voice with actually shows up as if as if it's any normal other orders. And so there is no difference for them in terms of operational workflow.

Brian Bell (12:10.961) Well, and I think another benefit to the restaurant too is instead of using something like DoorDash, which is kind of probably what they're they're currently using, or Uber Eats or something like that, you know, even DoorDash for a walk in order or a delivery order is gonna charge you twenty five, thirty percent, right?

Jay (12:26.892) Yeah. In fact, I think that was another reason why we actually sort of pivoted into more into the direct ordering and not just the voice bot. I know the voice bit name, kind of stuck to it, but the idea is that the idea about the door dash, I mean not specifically I'm gonna target Doodash here, but like any third party. is that during the COVID COVID years, they really got into because restaurant owners they just didn't have any drivers, they just didn't have any logistics to support their business, and third parties were the ones who were able to do that. And so what ended up happening is all these users, all the customers, their loyal customers, by the way, they are now on the on these apps and that's how they order. The issue with this is that they now have to pay any anything between like twenty to thirty percentages of the of the margins to these platforms and it's either customers or the owners have to bear it. And with VoiceBit now what happens is now you just call the restaurant. that was one of the reasons why all these customers were going into the third parties because they were not answering the phones, because they were busy running the kitchen. And now because the voice bit is able to handle any number of calls, they no longer have to go to go to these third parties. more importantly, we also pured into this web ordering. So we also now cover that. So you no longer have to lose margins into those websites as well. And that way we are trying to we are trying to basically make the make the restaurant owners own their own. own their customers again. That's what we are trying to do basically.

Brian Bell (14:28.743) So what's the pitch? Imagine, you know, we're in an elevator and I'm a pizza shop owner and you got thirty seconds to pitch me. What's the pitch?

Jay (14:35.436) Mm-hmm. Okay, let me think about this.

Brian Bell (14:41.245) Put you on the spot.

Jay (14:43.69) (14:49.398) Yeah. So the pitch is simple. We just we want you to own your own customers. One thing that a lot of people, a lot of restaurant owners in fact, might not know is that 70% of their business is being run by their core loyal customers. That is like 30%. You have to protect them. And there is no reason for them to go to this third parties or anywhere else and lose margin on it. You need to own them, you need to provide loyalty benefits, you need to provide personalized assistance, all those things. And so those are the customers who are probably running your business and you need to protect them. And so that's the pitch.

Brian Bell (15:31.247) And if you're s if you're sending them to some third party app like DoorDash, they're maybe thinking, there's like three other pizza places that I haven't tried that DoorDash is advertising because those other pizza pizza places can advertise on Do DoorDash as well and do a sponsored if somebody searches for pizza or somebody's even like making a pizza order to your shop, they could see a an ad for another pizza place.

Jay (15:38.924) Yep. Yep. (15:46.434) Yep, yep, yep. (15:55.181) Yeah, it's also in a destruction. Yep, for sure. Yep.

Brian Bell (15:57.446) Right. (16:00.881) So what's the how do you price it?

Jay (16:05.12) As in like okay, I need to ask like is it like you're asking the actual question of the pricing of the subscription?

Brian Bell (16:10.641) Yeah. Yeah. Yeah. For a pizza shop, like how do how do I get started? Let's say I want to do it, like how much does it cost?

Jay (16:17.114) yeah, so we we we have a couple of ways of pricing it and it's up to you as a biz business owner. we do it through usage based. So like if you want to just use it for the fake hours, let's say, you can just turn it on, then that's the time we will charge only the time where the AI speaks and takes an order. there is also a fixed subscription where you can just pay the fixed monthly subscription and then that will cover unlimited usage. And there is no commission involved here, there is no hidden fees, hidden charges involved here. Whatever you see is whatever you get. And so that's the basic idea here.

Brian Bell (16:58.877) So your team has talked about processing tens of thousands of orders, worth millions of dollars in just a few months. What's the metric you personally watch? more than that headline number.

Jay (17:10.14) I'm sorry, but on that one, I would change that question. I actually did, but I didn't share with you that one. So basically that was a long back. now I just wanted to say your team talked about placing order orders worth millions of dollars in just a few months because we are almost beyond that. And the another thing is as a team want to leak the even the matrix to the competitors at this point. so I'll just put Do we have a chat here? No, right?

Brian Bell (17:42.863) yeah. you can put it in the chat.

Jay (17:46.126) You'll see. (17:50.882) Yeah, I just put it there.

Brian Bell (18:05.979) Yeah. So what's a what what metrics do you like to to watch when you manage the business?

Jay (18:12.19) yeah, I think so the total order value number is a nice one. I mean it definitely tells us a lot in terms of the adoption. However, the number one metric that I personally watch for is the AI usage. and and the thing about that is we have been seeing constantly like the usage has been increasing, and that's good for us. I mean, we don't really care as much about I mean we do care about locations for sure, but then when the AI usage dips, what that tells me is the engagement rate is dropping. People the people now are starting to just drop or ask for the live agent. And so that's the biggest indicator for me personally. And that's the one that I always watch out for.

Brian Bell (19:04.605) Yeah, so it's actually calls successfully completed, orders successfully completed. Once they start talking to the the AI, it's again they get all the way through and make the order.

Jay (19:10.935) Yeah ex (19:14.934) Exactly, exactly the deflection rate has to be low. so that's one of the indicator and it's also the byproduct of it. So like if the AI usage itself drops, that anyways tells us that a lot of calls are basically people are asking for the live agent. Yeah. So yeah, that's the one we care a lot about. And the good thing about that is we on the on the AI usage itself, it's kind of funny because we saw a drop And we did not really lose any orders or the or the trend of the order was not dropping. And the reason it was dropped is because we introduced a memory layer in our agents. And what ended up happening was that a lot of these customers, you know, that I mean I share with you, right? Like they were like regular customers. For example, if you if you already know a pizza place, and then you usually call them, you already have your usuals. And we saw a lot of people were asking about it. And so we introduced this memory layer. Is now user have to just say yes, yes, yes, and yes. That's it. Because agent will already pump with hey by the way, hey Brian, thanks for calling. Would you like to have your usual large half an hour pepronium cheese? You say yes, yes, yes, yes, and that's it. That's the entire call. It's it's kind of wild because all you see is like four words from the customer and they actually placed an entire order. And so it was amazing.

Brian Bell (20:41.373) That's cool. what has been the biggest surprise? Cause you've been now applying AI out in the field for small businesses for at least a couple of years. Like you know, AI is moving super fast. I'm guessing, yeah, what's been your biggest surprise over the last couple of years of doing this?

Jay (21:02.828) I think initially I think l last year when we started this, it was comparatively hard to convince the customers that how this AI agent can actually help them, how much that it can help. I mean nowadays, I mean because of a lot of obviously the news media and all those things and the competitors as well, it it's it's it's not as hard, but I think a year ago it was actually very hard to c convince a restaurant owner that y they can actually benefit from it. Because if you think about this problem, it's actually invisible to them. Like they don't really know how many c orders they actually lose to this missed order, missed calls. They don't really know about this. It's until you show them. So what we had to do was that, you know what, let's just run it for a week or two weeks and see what

Brian Bell (21:59.88) No, do a pilot. Yeah.

Jay (22:01.334) Exactly, and see what the data says. And that's when when they say actually the order values, and sometimes more importantly the parallel orders, like during the peak times, two or three orders calls running at the same time, those are the orders you are anyways going to lose. Like there is no way you can even have those orders. you are anyways going to lose them. And so those are the orders and those were the data points that that convinced them. And so that was the hard part for us to even figure out how to even convince them.

Brian Bell (22:34.749) What's the typical order or I don't know, revenue uplift that you that customers see after they implement you guys?

Jay (22:43.948) In terms of the percentages, like yeah, I think they see anywhere up from like five percentage to eight percentage in terms of the in terms of the upsell revenue lift, which is pretty surprising because and it's like when you order a pizza, when you are in a rush, you might even forget about drinks or a dip or a side, something like that, right? And so if the age I mean agent has all the time in the world, right? So it can actually fetch all the contextual information about you, all the information about the inventory of the restaurant, all the knowledge about okay, what goes well with this dish, and then it can do the targeted upset. And actually that is one of the biggest reasons why a lot of restaurant owners are seeing a value in the EI agents. Because if you think about it, it's like if the if you hire a human if you have you have to train them with all of those things. not just your menu, but like how do you how do you make sure they upsell these things? most of the time the restaurant are not there even. Like sometimes they are they are there only on the Fridays, probably. And so the AI agent is basically an employee that is like the smartest employee, probably you can think about it, who can just who knows your menu. top to bottom it knows about the customers, their history, their preferences, and then it can just do the amazing upselling.

Brian Bell (24:21.255) So the website part is interesting because I mean I'm sure that existed out in the market, you know, fifteen years ago, right? but the voice the voice part is a new part.

Jay (24:23.746) Mm-hmm. (24:27.448) Yep, yep. Yeah, exactly. I think website so there are a couple of players there. I think owner.com, slice, those are the two things came to my mind. But yeah, they they were there e even even before like twenty twenty. So the AI is the new part. However, if you look at our websites that our restaurant owners use, they are they are not like a static website. They are actually interactive. You can actually talk to the website. If you talk to the website, it actually does things for you. So you don't have to really scroll through things. You just talk to the website and literally during in real time does the things. But yeah, I mean those are the things that that's the bar now. That's the that's the bar now. You no longer can get a with the static website.

Brian Bell (25:22.141) When you implement both the voice and the website, what percentage of the orders you handle flow through each channel?

Jay (25:30.412) Yeah, I think it's it's it's almost equal. And it's surprising. initially before going into the restaurant, because we don't have much con we didn't have any context in the restaurant space, we thought phone ordering would not be as big of a channel. And because you you hear a lot of people talk about it's a dying channel, it's it's a declining channel. However, we consistently see it's like It's like the same as websites, web orders. And so when we actually and the web orders was something we implemented afterwards. And when we did that, we at this point cover more than half the revenue for any restaurant that we are deployed at. so yeah, it's it's pretty significant actually.

Brian Bell (26:19.079) Yeah, so you're not you're not alone. you know, I think the press is already saying restaurant voice AI is a crowded market. Vox and Lohman both raised around the time that you raised. How do you kind of stay ahead of your competition and do you feel like it's a winner take most market? And yeah, what how do you think about competition?

Jay (26:40.876) I think I personally think it's it's a positive news to me. a couple of things. So the market is actually pretty huge. The restaurant industry is it's huge industry. And i i yeah, yeah, exactly. Yeah, exactly. It's it's a huge market. In fact and so I don't believe there is space only for one player or only one player can serve all the

Brian Bell (26:54.183) Yeah. How many millions of restaurants? There's millions of restaurants, right? And in the US alone. Yeah.

Jay (27:09.72) all the kind of restaurants and I I do believe a lot there is a need for a lot of products with different different capabilities. so yeah I don't see that it's a winner takes all market. we will definitely see even more in fact more player coming into this one as we see more and more the capabilities in improved for the voice in general because if you remember the voice was not such an obvious play. even a year ago. it was always about the chats, chat experiences, the coding and all those things, right? Voice was never believed to be the centerpiece or the sem center talk of the whole AI related companies but yeah I think in terms of how do we get ahead of it. So initially when we started the voice bit We in fact made this mistake of just building the voice voice bot. because that's the sexiest thing to do, honestly. the other parts are pretty boring and ugly, like the whole plumbing, the integrations and the kitchen operations, the the payment, the delivery dispatch, all those things, right? However, if you talk to any restaurant owner, they don't really like AI. I mean, they don't really want AI, not like so, but they don't really want any. What they want is a solution that just works, right? They're not like, you know what? I'm gonna just buy an EI service for my restaurant, it's gonna magically improve the operations. they just want to they just look at it as a business problem. And so what they want is my phone orders has to be automated. And for that to happen, you cannot just get away with building the voice bot. You have to do a lot of things. beyond voice agents. And that's how we just stay ahead. Like we we we we we develop. We take the first principle approach. We stay closer to the restaurant owners. They give us the feedback how they want to have the product, what more they would like to have, and that's how we just keep on improving.

Brian Bell (29:28.125) What are you working on now? What's on the roadmap for the next year?

Jay (29:33.027) I think our goal is to now that we have solved the voice player, our goal is to become that interaction layer for all the direct ordering channels for the restaurants. And that could be anything. It could be web ordering, it could be phone ordering, as you might know. We even recently launched Chat GPT ordering. So our our goal is to be this system where restaurant owners don't have to worry about their direct channels. They just rely on WisePit and it's infrastructure. and it will always be the commission free. I mean, okay, you know what? Mark that off. I don't know about that. So it will always remain that way. but

Brian Bell (30:24.295) Mm-hmm.

Jay (30:32.406) Our goal is to become this system of record or interaction layer for all the restaurants where all the direct ordering is done through us. And by that they can have a direct relationship with the customers. They can regain their customers from the third parties. They can have those margins back and then they can serve those regular ones, the loyal ones, with loyalties and marketing and everything. And so. Yeah, that's the goal.

Brian Bell (31:03.185) Yeah. Why why the chat GPT ordering integration?

Jay (31:08.364) So, quite frankly, I mean, I'm not claiming, we are not claiming that that's where the world is heading. However, we have seen a decent signal that people have started doing the shopping over Chat GPT. And so there is an enough demand, there is an enough signal for us to make sure the restaurant owners are ready for that. And that's how we communicate to the customers that Let's say tomorrow, in let's say a year or two years from now, if there's a boom on Chat GPT that I don't know, a younger generation probably would start doing ordering from there, then you are covered with us. You don't have to worry about that. So all of those aspects are covered. And so those are the reasons why we are basically trying to develop these technologies.

Brian Bell (32:02.107) Yeah, that makes sense. It's just another channel. Just like Google Maps, you you find the phone number, you call the restaurant, Chat G P T, you say, Hey, I'm looking for Burmese food. okay. Yeah, here's one nearby. I can order right through the app. Great. Okay, great.

Jay (32:12.428) Yep. Yep. (32:18.807) Yep, yep.

Brian Bell (32:20.337) What are you what are you excited about long term, like five, ten years out?

Jay (32:27.374) I would take a pause, I'll take a note. (32:43.19) Yeah, I don't have the note, but I would just try to s think about I mean five to ten years from now is kind of in the in the in the world of AI, it's quite a long time. It's kind of hard to imagine how How young even the Claude Code is. It's it's kind of wild. Like it's probably what, one one and a half years ago it was launched.

Brian Bell (33:12.285) Cloud code, yeah. Maybe a couple of years old, yeah, I'm not sure.

Jay (33:15.048) Yeah, it's it's it's it's not older than that. And it's you see the impact on on the on the software engineering that it has created. And so f five years from now, it's it's kind of wild to even imagine how this technologies will will have an impact on the restaurant industry. However, one thing I can definitely think about is margins are definitely going to get compressed. I think we I think that world of charging 30 percentage is probably gone. And if not already, then probably in the near future it will be gone. Because as more and more of these capabilities will develop, more and more of these apps will emerge, services will emerge, you will see this compression happening. And secondly, I think I will I am definitely seeing the productivity boost because of the AI. And that that actually increases the experience of the services, the quality of services. So for instance, if if you walk into the restaurant and you would not like a a a server just rushing through your order, giving you like I don't know if you are in in person rushing you an order just so that he has to take the call. I mean he can take a call, right? You would not like that. and so I I I do see those aspects. Will will get better, the quality of service, the availability in general and the margin compression. I think those are the those are the things that I see happening in near future.

Brian Bell (34:59.165) What's been the the biggest adjustment moving from, you know, a big tech company to being a startup founder?

Jay (35:07.394) Yeah, I think at big tech, I'm not sure y you're also from Microsoft. You worked at Microsoft before? Yeah, right? I think you might relate this, but a lot of things at big tech are being given to you. You assume certain stability. So you assume things to work certain ways. yeah, you have an expectation that

Brian Bell (35:14.843) Yeah, I was there for four years, yeah.

Jay (35:33.633) If I if I press this button for the deployment, it will actually go through all the clouds in the world. You have those expectations set up. Wherein w in startup, you have to actually build those things. And you have to make sure those things are there for you. So it's it's a different world out there in the startup altogether. However, for me it was not that hard honestly, because I always was in this trenches building things, tools. for myself on the sides. so I was alived it. But yeah, that is definitely a change. You can no longer assume that things would just work because somebody might have taken care of it. You have to be owner of everything.

Brian Bell (36:21.085) Yeah. Well, let's do some wrap up questions. kind of related to my last question is, you know, what's something you believed with total conviction in your Azure AI days at Microsoft that you'd now argue against at VoiceBit?

Jay (36:27.534) Mm-hmm. (36:36.536) I think one of the things at big tech or at least at Microsoft that I was conditioned to do was sort of making sure things worked end to end accurately and then ship it. I think in startup what you actually do is completely backwards. You you actually take a POC. You take it to the world and see how world reacts and then you work backwards from that and then you build your things. You it's a completely one eighty one eighty from what you do in the Microsoft. So that's the biggest shift that I had. Yeah.

Brian Bell (37:24.221) What did you used to be certain about that now through building voice bit you've slowly changed your mind on or you've er eroded the the certainty?

Jay (37:37.635) I think it's about the willingness of people to talk to or at least trust even an AI. I genuinely did not believe that voice ordering was such a big pain and in fact putting an AI agent on it would you even even remotely solve it? I did not believe that until I saw that. So when I saw the people actually order into it, that's when it just completely changed my mind. and then yeah, like people what they actually care most about is the quality of the service and availability of the service and not the actual like who is actually giving them the service. So that was the biggest, yeah, clarity.

Brian Bell (38:31.165) What's a popular idea in restaurant AI or restaurant tech right now that you think is wrong?

Jay (38:38.114) I think as I said earlier, restaurant owners don't really want NAI. They actually want a system that solves the business problem for them. They don't they don't want an AI just for the sake of it. if you if your if your solution doesn't really solve a problem end to end for them, then it's for no use for them. Yeah.

Brian Bell (39:00.145) No. What's the best advice that you've ev ever received?

Jay (39:08.334) Have to think about that. I actually thought about this a lot. (39:27.97) Yeah, I'm not able to think about do we have a way to change this quite I mean not even this one. couple of them because I actually thought about this a lot. (39:54.892) And you're asking what's the advice that I received at at Microsoft?

Brian Bell (39:59.678) Or just like generally, what's the best advice you've ever received?

Jay (40:04.801) yeah, I sure. I think (40:11.968) The assumption that you might have for the world might not necessarily be true. So it's it's it's an advice to to to be ready to be wrong, to be proven wrong when when presented with with a with a contrary reality of the world. and so just to be open, just to be curious. I think I I I like how Steve Jobs said that. They stay foolish. stay exactly. And so that's the that's the exactly.

Brian Bell (40:44.573) Stay hungry, stay foolish. Yeah. Yeah. So great, great speech. If anybody's listening, hasn't listened to the Steve Jobs commencement speech at Stanford. I think it was like two thousand seven ish, somewhere in there.

Jay (40:54.048) Is the best way. Yeah, it's an amazing switch. Yeah, it was Stanford, right? Yeah, it was an amazing switch.

Brian Bell (41:03.719) What a great yeah, what a great speech. what's a question you wish p more people asked you about VoiceBit?

Jay (41:12.803) I think It's it's not i i it's not about the okay, I have the note just a second. (42:11.022) I think one of the one of the one of the under underappreciated aspect about the voice agents and their impact on the businesses is their ability to actually talk in different languages. It's i i i if you think about it, like restaurant owners when they want to serve different age different different cultures, right? people from different cultures. Th they cannot like hire multilingual phone operator. Exactly.

Brian Bell (42:48.423) Right. People answer on the phone and speak Spanish just as well as they can speak English or indie or whatever. Yeah.

Jay (42:53.91) Exactly. And just just to add to it the accents, different accents. Like it's just insane like how hard it is to even understand those things. And so with AAI technologies, I think what it has enabled is more of those people or those customers are now able to access these services. And so it has actually made these services accessible to even broader audience. I think that's the best outcome. One of the best outcomes from this technology.

Brian Bell (43:27.453) From all your computer science education, 'cause you have a master's degree, what do you still use the most?

Jay (43:34.603) I think I still I I still am a fan of rigid data structure and algorithm. I I I I famously call have my name on my Discord server as a big older. I I really care still about the about the underlying structure of the of the systems because I do believe a good structures actually give you a good outcome, like overall in general I know in n nowadays the white coding world people usually think that you can just spin up the entire entire software with just a wish maybe we are still not there maybe in a in a near future we might be there but at least for now I still do believe the fundamentals of the computer science are still as relevant and I still use them a lot in my day to day.

Brian Bell (44:34.063) If voice bit, you know, exited or disappeared tomorrow, what's the first thing you would go go out and build next?

Jay (44:43.27) I think I have a very interesting answer. So I for a very long time I thought about building hedge fund that was driven by AI. I still am on that one. it's been a very, very long project and I've been trying to do that since several years at this point. I I have learned so much about different technical analysis, different techniques that this funds invol I know they do this sub millisecond trading. I'm not gonna reach there. But at least that's the one thing that's a that's the one thing that interests me the most, apart from voice bit of course. And that's the thing that I would probably just give it a shot.

Brian Bell (45:30.141) probably mess misremembering the details, but you know about this fiber optic line that goes from, you know, the financial district in New York to Chicago. And they like they spent like, I don't know, forty, fifty million dollars so they could just shave off like five milliseconds of latency or something like that between between the two exchanges.

Jay (45:38.775) Yeah, it's (45:44.885) Exactly. Exactly. And I didn't realize I think it's what, ninety percentage of the trades are happening through the algorithmic trading? It's kind of wild already. Yeah, it's it's it's kind of wild, yeah.

Brian Bell (45:55.617) yeah. Yeah, it's the majority, yeah. (46:00.99) So where can folks find you online if you're obviously if you're a restaurant owner and you want to get in touch, or maybe if you're a VC's listening wants to get get in touch with you, what's the best way to get in touch with you?

Jay (46:10.188) Yeah. Yeah, we share we share our I mean we have our own LinkedIn page. You can follow us there for the latest news. I have my own page there. I regularly share articles, the my opinions on the AI trends there. so you can follow me there. and then of course our webs voice bit.ai website. we regularly share new talk new blogs, the the interviews with the restaurant owners, our expos, everything.

Brian Bell (46:40.839) W one more qu bonus question, because this is very in the zeitgeist right now. What's your what's your P doom for for AI?

Jay (46:54.122) I think with the latest so I I forgot the name of the problem that they solved, but the nearest yeah, it's yeah, exactly. It's

Brian Bell (47:00.913) The Navier Stokes conjecture, yeah.

Jay (47:07.936) I I I I I I have about ten percentage, P ten I would say. it's

Brian Bell (47:12.978) Wow.

Jay (47:16.694) Yeah, it's it's it's not looking good.

Brian Bell (47:19.293) I'm I'm a lot more optimistic than you because I I feel like we'll have lots of super intelligent AIs. and we'll have a sus a whole society of super intelligence and they'll be collaborating and competing. And you know, I think also we'll start augmenting our own intelligence with AI soon, with neural interface interface brain PCIs and stuff like that. So

Jay (47:46.882) Yep, yep.

Brian Bell (47:47.964) I think I think it's more probable probable that we merge with you know super intelligence in the future and we become a new a new species, right? And and some humans will just choose to remain human, biological humans. And but I I I would I pr I would predict over the next decade that we become we all become super intelligent.

Jay (47:55.383) I I see. (48:06.702) Mm-hmm. (48:11.554) Yeah, I mean, of course. I think it's just that the the the capability of this technology is just and the way it's the the speed at which it's developing, it's just almost fire.

Brian Bell (48:22.907) Yeah, we seem to be in a hard takeoff. Yeah, we seem to be in a hard takeoff scenario.

Jay (48:27.438) Yeah, it is almost frightening.

Brian Bell (48:30.311) We're right on c we're right on the AI twenty twenty seven paper timeline. If you read that. Yeah, we're right like right on that curve. So yeah, right on time. It's like, ugh. So hopefully we can get the alignment stuff figured out.

Jay (48:34.85) Yes, yes, exactly. Right on time. Right on time, yeah. (48:43.404) Yep, yep, exactly. But yeah, it is I yeah, it is kind of tricky. And I I do worry about it. I actually talk to my co founders regularly on that one. Like like every time when I see these new models and new new news coming up, new developments. It's just it just worries me that you probably have to just do things faster than what we might have thought about earlier. But yeah, I mean it's just yeah. Let's see how it goes. As you said, we have survived a lot. We have survived what? Like the I don't know, the the worst of the worst viruses, the the disease. The exactly. All of that. So yeah, I mean and it's been like what?

Brian Bell (49:14.897) Yep. I hope for the best. (49:28.731) Yeah. Pandemics, wars, the threat of nuclear war, all of that. Yeah. I'm pretty confident we'll we'll we'll get through it. and somebody made a point, I think it was on the Moonshots podcast, it was Alexander Wisner Gross, one of the best thinkers on AI. He says that if AI, if there's super intelligent AI out there and it destroys, you know, everything, all you you would look around the galaxy and it would turn would have already turned everything into computronium or or paper clips or whatever.

Jay (50:04.598) Yeah, I mean it's and also I'm I'm not sure if you know about this paradox. If if the if the species if there is a species who's actually smart enough, it would actually be invisible to us. I already. Like if if they are already smart enough, they would be invi invisible to us at this point. And

Brian Bell (50:19.515) Right. Yeah. (50:24.701) Yeah, they'll have technology that's just like magic to us, like holodex and cloaking technologies and all kinds of stuff. They're probably not even using the same radio waves as us. They're using some sort of sub space Star Trek stuff that we can't even detect, you know?

Jay (50:27.659) Zarry. (50:31.169) Exactly. (50:39.426) Yeah, I mean that's where that's where the the whole idea the I'm not sure if you know I I'm not sure if you what's your opinion on P equals to N P like T versus N P. Like do you think do you think it's

Brian Bell (50:40.711) Yeah. (50:50.957) All right. No, I don't I don't I know of it. I I don't know much about it. May maybe tell the audience what that is.

Jay (50:58.22) Yeah, I mean the P P versus NP is basically the any problem that you can verify are basically the problems has the polynomial time of complexity. And so we don't really know whether all the problems in the world that you can verify easily are also can be solved easily. In easily means here the probabil polynomial time. We don't really know that. But in history we have seen, like for example, the primary check. it it it used to be believed of non NP problem basically, non-deterministic polynomial type. But then until I think recently, probably in two thousands, two thousand two, the AKS algorithm came up and they actually broke through the wall and found out the polynomial type. solve that. And so it's it's it's it's it's kind of weird because if using during I mean that's one of the millennium problem by the way, one million dollar problem. if they are able to solve that we'll see how it goes. I don't know like how the whole entire encryption and everything is on that assumption that P is equal to P is not equal to MP. but it's kind of wild if we can do that, then a lot of things has to change.

Brian Bell (52:16.285) Right. Yeah.

Jay (52:26.476) because of just that belief. but yeah, we are definitely an interesting time.

Brian Bell (52:33.415) Yeah. Well I really enjoyed it, Jay. Great to catch up.

Jay (52:36.867) Likewise, thank you.

This article is for general informational purposes only and does not constitute investment, legal, tax, or accounting advice, nor an offer or solicitation to buy or sell any security or investment product. Investing involves substantial risk, including possible loss of principal, and past performance is not indicative of future results. Full disclaimer.