Latest / State of the AI Union / Nitya Arora: Product & Tech Tidbits, right at your Seller's Fingertips
Transcript
- 0:06Welcome to this week's episode of the State of the AI Union.
- 0:09And I am your host, Laura. Today I'm joined by Nitya Arora,
- 0:13cofounder of Pepper AI. And Nitya represents a new
- 0:17generation of founders who are truly AI native.
- 0:21She entered the workforce as Gen.
- 0:23AI was becoming real, not just the add on.
- 0:26And she's already built and sold a company.
- 0:29So now bring a new kind of company to the market.
- 0:31Pepper, I wanted to give you a chance that you have to
- 0:34introduce yourself and also Pepper, tell us about both of
- 0:38both of you. Yeah, of course.
- 0:40So Pepper is a real time copilot for sellers that allows them to
- 0:46understand all the technical context and all the product
- 0:50knowledge in a company by instantly surfacing it while
- 0:54they're on call, along with a post called Knowledge Routing
- 0:58system that makes it much easier to understand organizational
- 1:01intelligence. So the idea comes from kind of a
- 1:06long background of working in enterprise for a short amount of
- 1:11time, but feeling a lot of pain points.
- 1:14So I, I started working at NVIDIA about four, I'd say it
- 1:21was about like a year ago. And I'd been working there for
- 1:24four months. And what I experienced was all
- 1:27the knowledge in that company was very siloed.
- 1:29People worked on very individual projects.
- 1:32You know, everyone was insanely intelligent, but they were all
- 1:36working on very specialized things.
- 1:38So anytime you had a question, you had to go to that person who
- 1:41owned that project, who'd been working on it for two or three
- 1:44or four years and asked them everything about it.
- 1:46So there's these people getting pinged for knowledge every 40,
- 1:50you know, minutes, just people asking them, hey, what did you
- 1:52do with this line? Or like, why did you push this
- 1:54code? And so we took that pain and we
- 1:58realized all of this knowledge lives in Slack, it lives in
- 2:01Notion, it lives where people are talking.
- 2:03And that's where we decided that it would be a very interesting
- 2:07idea to have enterprise search across the board.
- 2:10That was very privacy preserving.
- 2:12And I understood all of the knowledge in a company.
- 2:15And as we roll this out to engineering teams, we saw this
- 2:19problem with sellers facing the exact same issue, but not having
- 2:23a conducive interface for the place where they live, which is
- 2:26in calls. And so seeing a lot of
- 2:30conversational intelligence today, we were like, what
- 2:33differentiates helping the seller Wilder with the customer
- 2:37versus post call where you've already lost them, where 10% of
- 2:41their attention span is with you in their inbox?
- 2:43How do we get that knowledge into the call?
- 2:47And that's exactly how Pepper came alive.
- 2:50OK, so niche, it sounds like you originally built the product to
- 2:53support these engineering teams that were looking for like
- 2:56information on like how to build better ship faster, that kind of
- 2:59thing, getting institutional knowledge about the code and and
- 3:03the product. And then you decided to pivot to
- 3:07sellers because you also noticed that sellers were having these
- 3:10challenges when they were talking to customers, Like
- 3:15what's the, what's, what's the, what was the solution for
- 3:18sellers today that you were seeing and how is that impacting
- 3:22their work? Yeah.
- 3:24So the solution that we've seen is a very conducive UX to being
- 3:30on their screen and away from it so that they can choose to
- 3:35toggle it on and off. I think the biggest part of what
- 3:37we're doing is experience. It's that you don't want to
- 3:40invade a call you don't want to be.
- 3:42I mean, bots are OK. I think people have have grown
- 3:45accustomed to them, but we don't even take the bot approach
- 3:48because even that feels a little like a stranger watching you in
- 3:52a room. A lot of settings that you know
- 3:55sellers need to work in, they're educationally restricted or
- 3:58governmentally restricted, so you're not allowed to record
- 4:01calls and you still need a lot of this information.
- 4:03So we operate on the real time and optionally as opposed to
- 4:07call analytic tool because there's a lot of those out
- 4:09there, but there isn't a lot that sits on your screen as a
- 4:12transparent overlay that you can switch on and off that surfaces
- 4:15answers when you need them. And that also can help you with
- 4:20OK, now that this call is over, these were the 10 things that we
- 4:24talked about. These are the action items that
- 4:26we can click buttons for right now.
- 4:29So imagine, you know, you're working on, you're working with
- 4:32a customer and they say, OK, here, let's follow up in
- 4:35February. Usually a lot of these calls
- 4:37have 27 different action items. So let's follow up in February.
- 4:40I'm going to loop in my technical team, I'm going to
- 4:42loop in, you know, Jane from like delivery.
- 4:46And I'm going to end up, we're going to, you know, talk about
- 4:50some of the points that we had today, some of the technical
- 4:52projects you worked on in the past.
- 4:54All of these are trains of thought and more knowledge that
- 4:57you need to follow up with after a call.
- 4:59So Pepper aligns all of that as soon as it happens, and then it
- 5:03helps you route that knowledge to the people that can help you
- 5:06with that. Half of a seller's job is
- 5:09internal collaboration. It's understanding how do you
- 5:12position everything in your company so that you can deliver
- 5:15to the customer in the right way.
- 5:17And that's the part that Pepper helps automate and make sure
- 5:20that you do right as well. I actually have like 2 follow up
- 5:24questions on that. So I think the first is, you
- 5:27know, it does sound like it's a lot of a lot of other tools and
- 5:34systems are doing somewhat something similar, including for
- 5:37example, the post call analytics, following up
- 5:39summaries, making sure that we're capturing all these
- 5:42meaning transcripts. Tell me, tell me why those are
- 5:47not as desirable as this sort of like copilot approach.
- 5:53And then sort of like next question to that is, are there
- 5:59if the call pilot doesn't actually join the call, are
- 6:01there any kind of like permissioning and like sort of
- 6:05data privacy things that we have to be careful about on that
- 6:09side? Yeah.
- 6:11So on the side of the first question, I think the real
- 6:15utility of a real time assistant comes from being able to
- 6:19highlight exactly what you need to say on the call or bring up
- 6:24those concepts and contexts that you're probably missing
- 6:27otherwise. And this means like, let's say a
- 6:28customer starts talking about another customer they've worked
- 6:32with in the past or some company that they've worked with in the
- 6:34past and you didn't know it, but your company is like an
- 6:38extensively huge partner with them and like, you know, they've
- 6:42been working with them for a year.
- 6:43That is a great thing to mention, right?
- 6:46That adds so much credibility in the moment that builds
- 6:48relationships. And all that happens because you
- 6:51can't memorize a CRM of relationships your company's had
- 6:54because there's, you know, hundreds of sellers that are
- 6:56doing the same thing across your company.
- 6:58And maybe someone talked to someone from, you know, like
- 7:01Core Weave last week and they've been, you know, collaborating
- 7:05with them. Well, hey, I want that type of,
- 7:08you know, connection to be established on my call as well.
- 7:11And where is that information? I'm only going to get it if I
- 7:13talk to that seller or if I see it on that CRM.
- 7:15And those are the types of, you know, institutional pieces of
- 7:19information, which if we had a perfect AI sales agent that
- 7:22could take a call, which in my opinion isn't going to happen.
- 7:25I don't think sales is a very like automatable process, like
- 7:29to the point where you take a human out of it.
- 7:31But if we were to have one, it would be able to pick up on
- 7:34that. And our job is how do we assist
- 7:36someone in real time to have that type of information to be
- 7:39able to develop those types of connections without, you know,
- 7:42looking clunky and like being like, hey, you know, it might
- 7:45not even occur to them to even search for the fact that that
- 7:47person or that company was was a customer.
- 7:50And those are the type of things our copilot can pick up on and
- 7:53reference immediately for you to see if it's verified or not, if
- 7:56it's approved answer or not. And that comes into what we do
- 8:00differently post call is a lot of sellers today don't know what
- 8:05they don't know. And that's a perfectly perfect
- 8:09and normal problem to have. And if the sellers themselves
- 8:13don't know what they don't know, and, you know, there's a lot of
- 8:16conversational intelligence tools that, you know, help
- 8:19revenue enablement folks and sales enablement folks look back
- 8:22on calls and see, hey, you know, this is what went wrong and this
- 8:25is what went right. But a lot of it requires
- 8:27actively listening to the call and being like, hey, what do my
- 8:30reps know? What do they not know?
- 8:32Here are the summaries of their calls.
- 8:34Pepper is specialized in picking up those technical moments,
- 8:37those moments where you're like, hey, I don't know, I'll, I'll
- 8:42get back to you. And it picks up on that.
- 8:44And it's like, OK, now this is a technical moment that you like
- 8:48kind of didn't understand. Here is all of the knowledge in
- 8:51the company automatically given to you to be able to review so
- 8:55that you know this for the next time.
- 8:57That's a hard problem because that means you're doing
- 8:59enterprise search on voice data and making sure that that
- 9:03specific piece of information they're enabled on properly.
- 9:06And you're also giving that feedback to enablers where
- 9:08you're like, look, your team is really confused about this new
- 9:11product. They won't have any information
- 9:14about it. This is what you should be
- 9:16asking product and this is how you can route it within 10
- 9:19minutes if they answer all the questions and the time.
- 9:23So if I'm hearing you correctly then like the main
- 9:25differentiator that Pepper has over kind of a lot of the CI
- 9:31tools out there is that, you know, it's actually surfacing
- 9:34the product and technical contextual knowledge about the,
- 9:39the, the company or the organization to the salesperson.
- 9:41Whereas I would say, you know, some of the other tools may be
- 9:45more focused around the actual selling methodology.
- 9:48Do we have access to power? Do we have a champion?
- 9:50That kind of thing? Exactly.
- 9:52I think they're very sales driven and that's great.
- 9:55But there's a lot of companies where being sales and like, you
- 9:59know, kind of addressing A playbook isn't enough.
- 10:03You need to be able to go the extra mile to understand some of
- 10:07the technical concepts, be able to discover a little more on the
- 10:09initial calls so that you can pass something that's that's
- 10:12worth it to the delivery and technical team.
- 10:15Yeah. And and you know, actually let's
- 10:19go back to the the second part of the question that I had
- 10:21before, which was around, sorry, I forgot the, the, the privacy
- 10:26and sort of data restrictions here.
- 10:27Tell me how you're you're managing that.
- 10:29Yeah. So what we do is at the
- 10:32beginning of every call, especially if you're in a
- 10:34regulated industry, I think it's standard practice in general
- 10:37for, for companies to do this. But what we do is we like we'll
- 10:40ask like is it OK for us to record this call or start
- 10:43recording it for you? Like make sure that that kind of
- 10:46consent you have that from your user.
- 10:48So the onus falls on the company for like recording something
- 10:52they shouldn't be if if that's what they do.
- 10:55But calls are not automatically recorded.
- 10:57So that's a very important piece of this.
- 10:59It's we don't record calls, but we're able to get all of the
- 11:02insights even if you don't record the calls, that's
- 11:04another, that's a tertiary option.
- 11:08The idea behind it is we're able to help in real time either way,
- 11:14regardless of whether we record the call or not.
- 11:16And that's a very big unlock for a lot of different teams because
- 11:21a lot of them in government, in education, in these, when
- 11:25they're talking to regulated or more secure customers need to
- 11:28have some log of what happened, right?
- 11:31But they're not going to be clunkily taking notes away.
- 11:34They might want a summary of the conversation that's anonymized
- 11:37and we're able to provide all of that by the track or extracting
- 11:41details that don't matter to kind of the specifics of the
- 11:44conversation. But due to, you know, follow-ups
- 11:48and tangibles like technical questions and be able to route
- 11:51them based on their consent. So there's a lot of optionality
- 11:56built into whether you want to record a call, whether you don't
- 11:59want to record a call, how you're going to extract
- 12:01information from a post call without, you know, extracting
- 12:06the details that they don't want to be extracted, and that's
- 12:10that. But that's built into trying to
- 12:12make it as useful as possible. Got it.
- 12:15So the, the, it's the same functionality right now where we
- 12:19record a call and we ask them for consent.
- 12:21It's the same sort of consent sort of mechanism being sort of
- 12:24except, except that you guys are not joining as a bot, you're
- 12:27sort of operating the background.
- 12:29And that also gives you the ability to kind of be an AI
- 12:34assistant note taker if it isn't recorded.
- 12:37Yeah, exactly. Awesome, awesome.
- 12:40Now tell us a little bit about the tech, right.
- 12:42How does this, how does, how, how are you able to like bring
- 12:47everything in and, and, and how is it so special?
- 12:49How are you guys doing it differently?
- 12:51Yeah. So I think what what really
- 12:53filtered into our experience was actually when we were working
- 12:56with engineering teams, which we still do with our original
- 12:59product that helps users get information fast from all of
- 13:05their sources of truth. And that integration pipeline
- 13:09really helped us extract a lot of information in what we
- 13:14thought was a fast way. But with voice, you're working
- 13:17with another layer, which is that like latency of the seller
- 13:20being able to see the information fast and see it in a
- 13:22way that's digestible. So those are the 2 main
- 13:24technical problems we work on. The 1st is the latency, and the
- 13:27second is being able to see that information in a way that's
- 13:30conducive to their personalized tone of voice in a way that they
- 13:34would say it and then be able to go down deeper for technical
- 13:37details. Now the idea behind Pepper is
- 13:39it's not supposed to be something distracting.
- 13:41It's not supposed to be something that takes your
- 13:42attention way. It's supposed to be you take a
- 13:45look at it. It's easy to say, it's easy to
- 13:47understand, easy for you to be able to communicate and then
- 13:50move on or take some of the details, copy the answer,
- 13:54schedule a follow up, write the details in there.
- 13:57You know, have you know the 10 questions that they asked routed
- 14:00and then pulled back into your follow up emails automatically.
- 14:04But in order to do all of that, you need to be very personalized
- 14:08to what the seller talks about and you I never recall and you
- 14:12need to understand that. And that is a huge kind of
- 14:18technical problem for us, along with the latency, which is
- 14:21something we work on every day. So how can we get answers to
- 14:24come up in 500 milliseconds instead of 800?
- 14:28How do we get them to stream in a way that looks nice and a way
- 14:31that is like easy on the eyes and stuff?
- 14:34So all of those like very, very small things are a huge part of
- 14:37what we're doing technically because we kind of solve the
- 14:39search problem for ourselves with connecting to all of those
- 14:43different tools. It's more about how do we put it
- 14:46into this UX? Yeah.
- 14:48And and on the on the latency, is this something that you want
- 14:52to tell us about on how you're solving for that so that you're
- 14:54actually becoming faster? Yeah.
- 14:56So there's a lot of, I mean, canonically there's like a lot
- 15:00of chunking that people will do in different ways.
- 15:04We need to be able to roll call summaries UPS faster.
- 15:08So you need to be able to understand like, hey, what was
- 15:10the context of the past two minutes of this call?
- 15:14And how does that filter into what question they're asking
- 15:18right now? Another huge problem is how do
- 15:20we know when to show something right?
- 15:22So that also filters into the latency because it's kind of
- 15:26like a, it's, it's like a give and take relationship where if
- 15:29you have like intense actionability all the time, then
- 15:33you're going to ramp up your latency because you're going to
- 15:35be making a lot more calls. And so keeping a fine balance
- 15:39between that based on kind of and, and what we do a lot for
- 15:42that is a lot of user base prompting and it's part of the
- 15:44personalization and post call feedback is where are they
- 15:47getting stuck? And can we optimize how
- 15:50actionable our agent is for this specific person while they're on
- 15:53call based on when they're getting stuck?
- 15:57So can we make fewer calls? Can we make sure that there are
- 16:02less people that, you know, we're pinging post call to make
- 16:05it easier on the system? How do we, you know, optimize
- 16:09how much we're using it so that we don't we can get even faster
- 16:14results on a personalized level? And a lot of that also comes
- 16:17into caching. So for users, like for specific
- 16:20users, if we know that they're being asked the same technical
- 16:22questions, we can actually cache all of that in a way that is
- 16:25easy for them to access and retrieve while they're using the
- 16:28Copilot locally. Got it, got it.
- 16:30And then tell us about the connectors, Nyja.
- 16:33Those are the most important thing, right?
- 16:35Cuz you're basically connecting all of these systems.
- 16:37What are some unique ways that you're connecting to the data
- 16:41that make it easy for you to pull everything?
- 16:45And then what about permissioning and data
- 16:47governance, that kind of thing? Yeah.
- 16:49So that was a huge friction that we anticipated when we were
- 16:53going to work with engineering teams in the first places, being
- 16:56able to have a secure way of connecting with all of those,
- 16:59all that different knowledge, especially with, you know,
- 17:01CSO's. So we came up with a very
- 17:05privacy protecting way of doing it.
- 17:08We worked with Open FGA and that kind of centered a lot of our
- 17:13approach to authorization and access to knowledge.
- 17:17And so we were able to model a lot of the relationships and,
- 17:22and make sure that RBAC is intact across all those
- 17:24different sources. Very, very method in a very, in
- 17:29a way that was very methodic. So we were really, really
- 17:31intentional about how we were, how, you know, a manager's
- 17:34information was displayed to an employee and making sure that,
- 17:38you know, let's say managers were talking about pay stubs,
- 17:41not that, you know, that information isn't what's going
- 17:44to be displayed to the employee when they're using enterprise
- 17:46search and it's not in the bounds of what they're looking
- 17:48at. So we need to be very rigid
- 17:50about that across all the integrations.
- 17:51We're always shipping integrations.
- 17:53To be honest, there was no outsourced pipeline that we
- 17:55could use that allowed us to be able to connect with all these
- 17:59sources in a way that was, you know, flexible enough to work
- 18:03with our customers and also promised all the security that
- 18:07we were talking about and the quality of search.
- 18:09And so those were the three things that we really focused on
- 18:12with connectors and and we had to build it all in house.
- 18:15OK, OK. Wow.
- 18:16Sounds like a big challenge. How long did it take you to sort
- 18:19of build that infrastructure before you could put the UI on
- 18:22top of it? Honestly, to build it, well, I
- 18:25think it took us around 5 months and I was like 5 of us, yeah.
- 18:30So it was definitely a long time.
- 18:32And I think it's also just kind of an ongoing process.
- 18:35It doesn't really stop. And that's part of it.
- 18:38That's part of why it's such a big challenge is because things
- 18:40are always changing across all these connectors.
- 18:42There's always legislate. Like I think four months ago,
- 18:45Slack decided like, we're not going to, you know, share all of
- 18:48this data outside of Slack because Slack AI like it's very
- 18:51useful for that. And so that's also another
- 18:54challenge we to deal with is like rate limits on that, being
- 18:56able to still extract all that information without hitting
- 19:00those roadblocks. Yeah.
- 19:02So, Nitya, this is actually the second time that you've started
- 19:05a company, right? The first company you had, you
- 19:07sold. Yeah, what?
- 19:08Do you think was and that was also, that was also recent like
- 19:11if I remember 2021, Yeah, yeah. Yeah, yeah, OK.
- 19:15So as a recent company you ended up selling it.
- 19:17What do you think is different about building a company this
- 19:22time? Yeah, I was, I was a kid when I
- 19:25started that coming. Like I didn't, I didn't know
- 19:26what I was doing. I just kind of, I, I think I
- 19:28started in my junior year of high school just because I
- 19:31needed to like build something interesting after I watched The
- 19:34Social Dilemma on Netflix in my class and it was around like
- 19:39data sharing and being able to I, I, I, what I watched on like
- 19:43Instagram or about Instagram on Netflix made me kind of like
- 19:48annoyed. I think I at the time, like they
- 19:51were talking a lot about, you know, people losing attention
- 19:54spans and yes, their data to companies.
- 19:58And I think that as like, oh, people can make money off of
- 20:01their data. And I was a very, very naive
- 20:04like like junior. And so I went on this like weird
- 20:09journey to make like a data selling app, but that's not what
- 20:12it ended up being. What it was more along the lines
- 20:14of was a, so I didn't know how to build.
- 20:17I spent like a year trying to build that app.
- 20:18I didn't know how to work React Native, which is the platform I
- 20:22started to use. So I started to learn that over
- 20:24the summer. And as I built it, I, you know,
- 20:28started to, I took a really long time to build it.
- 20:30I was a really bad app, like I think it was like 8 months in
- 20:33and it was so bad. But I started to make it better
- 20:37over time. I started to learn a lot of what
- 20:39made me like a technical builder.
- 20:41And it was great because I feel like if I didn't have that
- 20:43first, like, oh you know, I can just learn how to code because I
- 20:46want to make something for the sake of making it.
- 20:48I would have never actually thought I could be a founder.
- 20:50And I didn't know it was like the barrier to entry was that
- 20:52low. Wanting to build something of
- 20:55your own. And as soon as I, you know, it,
- 20:58it pivoted into that was like the first time I had to pivot.
- 21:02I was going crazy about it. Like I was like, Oh my God, I'm
- 21:05like a year into building this app and like, I don't know if
- 21:07you can actually sell people's data or like if people can
- 21:10actually like learn from this, like in the way that I want them
- 21:12to. So I need to move towards
- 21:15restaurants and like these B2C businesses that kind of had a
- 21:20lot of customer loyalty that they wanted to help with.
- 21:24Like most restaurants don't know who they're like top 20
- 21:27customers are coming in. Anyone know that those are the
- 21:30same people that come in for like a drink every Friday And
- 21:33they're the same people that live 10 minutes away and go
- 21:38grocery shopping at like Walmart every other Sunday.
- 21:41And all that information and data is really helpful to be
- 21:44able to bring customers back, you know, give them a voucher,
- 21:47give them some form of reason to come back even more, build a
- 21:52relationship with them and they just don't have the data behind
- 21:55it to crunch those numbers. And that's what we did.
- 21:57So we were there data middleman for customer loyalty, kept all
- 22:01of the data on the user side secure.
- 22:03And then we were able to build those relationships and
- 22:05marketing campaigns for them automatically with offers to
- 22:08suggest that they could put out on the app that people could
- 22:11redeem and then build a more organic relationship with them.
- 22:16Wow, that's amazing. I love your story about how you
- 22:19started something. And the common thread that I see
- 22:22between your lost company and your last business idea in here
- 22:26is just basically that data. And that data is just so
- 22:29important, whether it's human data, but it's also giving,
- 22:32giving both consumers and organizations a better way to
- 22:38serve each other. Exactly.
- 22:40Yeah, I think I was always interested in very external type
- 22:47connections, I guess customers with their users better that
- 22:51that sort of thing. And I had a whole fad with like
- 22:54marketplaces as well. Like I I loved every marketplace
- 22:57idea for a while. Yeah, Yeah, I love it.
- 23:00OK. So going back to going back to
- 23:03Pepper for a second today, when you are talking to customers and
- 23:07you're talking to them about, you know, this challenge that
- 23:11the sellers have and not having enough like product and
- 23:14technical information that they need.
- 23:17What are some key impact metrics that you're seeing your, your
- 23:25customers, you know, sort of surface like is it just it's,
- 23:29it's taking too long, the sales cycle dies.
- 23:32Tell us a little bit about that. Yeah.
- 23:34So I think that's you're exactly on point with that.
- 23:37Like it's taking too long and the, the sales cycle dies and
- 23:41another like more impactful metric, I think that we found is
- 23:46just the time to productivity for sellers, right?
- 23:50So we didn't think we'd be helping with enablement teams.
- 23:53Like we kind of went to account executives and we're like, hey,
- 23:56you know, this is a copilot that could help you.
- 23:58And we didn't know there was a whole like, I don't come from a
- 24:02revenue background, so I didn't know how complex, you know,
- 24:06enterprise sales could be. But there's an STR, there's an
- 24:09AE, there's an SE, there's an SA, there is like a CSM and
- 24:13they're all having to work with each other across the entire our
- 24:16revenue pipeline. And where I talk to like our
- 24:19customers, the more I understood that that's like those are
- 24:22relationships and information that has to bump between all
- 24:25four of these people throughout the process.
- 24:27You know, at some point, like the STR brings the lead and they
- 24:29know like what they brought them in for or how they brought them
- 24:33in, they need to be able to communicate that to the AE so
- 24:36that the AE knows like, hey, this is like the context I'm
- 24:39supposed to be talking to them about.
- 24:41And then the SE and the SA have what I'd say is one of the
- 24:45bigger jumps in this entire cycle of understanding all of
- 24:47that information and then being able to technically deliver it.
- 24:51And I think that that friction is where we're seeing enablement
- 24:55happen. It's where the time to
- 24:57productivity of a seller, a new person that's coming in.
- 25:00And fortunately, a lot of the companies we've worked with,
- 25:01we've been able to see these patterns because they have new
- 25:03sellers coming in and they're onboarding.
- 25:05A lot of people post acquisitions, post just a hiring
- 25:10cycle. And because they're doing, they
- 25:11have to onboard onto these new products.
- 25:13And the way to get knowledge from the solutions engineers
- 25:18throughout putting them on a discovery call, for example, is
- 25:21much better routed through us. And that metric we're seeing is
- 25:24like, OK, now the solutions engineer has to sit on 10 less
- 25:28calls every week because, you know, not being asked the same
- 25:3220 questions that they are on like every 14 calls they join.
- 25:36So that is a metric where you're saving time on the end of the
- 25:39solutions engineer and the sales engineer, which the SC team
- 25:42loves. And then for the seller, you're
- 25:45also enabling them because now they're learning the answers to
- 25:48those questions that they would have typically, you know,
- 25:52completely relied on the Southeast to carry the
- 25:55conversation for. Yeah.
- 25:58And I don't know if it's something on your road map, but
- 26:02as you're talking a little bit more about enablement and time
- 26:05to productivity, I was just thinking more about sort of role
- 26:09role plays and things before the actual live call.
- 26:13Or is there anything that can help sellers there sort of get
- 26:16up to speed before they get in front of the customer?
- 26:19Yeah. So we actually do a lot of pre
- 26:21call prep and a lot of it is based on like our own internal
- 26:25workflow for it just cuz we like, you know, I think we dog
- 26:28food what we use a lot And because of that, we're able to
- 26:32like iterate on it fast. But one thing that we were able
- 26:36to introduce is more than role play of being able to get all of
- 26:41because we're so linked into all the different sources in a
- 26:43company, being able to get all of the signals of that
- 26:46information and kind of build what I'd say are very custom
- 26:50talk tracks because of it. So be able to make those
- 26:53relationships like, hey, you know, we looked through this
- 26:55guy's LinkedIn and this customer, I mean, they've been,
- 26:58they've been talking to and posting about this other
- 27:00customer, like their company account has been talking about a
- 27:03recent partnership with this other customer that's also a
- 27:05shared customer that we have. And then being able to link that
- 27:08information into all the past calls that, you know, they might
- 27:11have had with other people from that company.
- 27:13And you know, even if this is a discovery call, being able to
- 27:17link in technical information that you should know before the
- 27:20call, just guessing at use cases and how this company is similar
- 27:24to other customers that you've worked with.
- 27:26So a lot of knowledge and signaling from what exists in
- 27:29your CRM and connecting that with what exists with, you know,
- 27:32maybe the calls that your CSMS have had with other similar
- 27:35customers that you can link back as issues that you can talk
- 27:39about have already been solved, for example.
- 27:41So all of that, imagine all of that predictive analysis like
- 27:46pre call because we're so integrated with all those
- 27:48different schools. Got it, got it.
- 27:51OK, awesome. Yes, yeah, OK.
- 27:53I wanted to pivot a little bit and talk about your background
- 27:57as a founder. I know that, you know, it
- 27:59discussed you had grown up in Singapore for about 9 years.
- 28:02You went to the Singapore American School.
- 28:04Tell us what that experience was like and what that gave you in
- 28:10terms of being an entrepreneur, setting up your own companies.
- 28:14What was it like for you? Yeah, I think entrepreneurship
- 28:18was like pretty drilled into me at SAS and a lot of it had to do
- 28:22with, I mean a like I've just seen a lot of founders in my
- 28:27life. Fortunately, I think that had to
- 28:29a lot to do with it. Like, I don't think looking
- 28:31back, I don't think I would have been a founder if I hadn't seen
- 28:33as many founders. And I guess that has a lot to do
- 28:36with how you grow up and like psychology, But that is
- 28:40something I look back at and look like, you know, if I
- 28:43didn't, if I didn't see those people, I probably would have
- 28:45been influenced to do it. It seems like a random decision.
- 28:48The second thing was at SES, like I think I had a lot a like
- 28:52they had a lot of extracurricular activity stuff
- 28:55going on. So there was like robotics
- 28:57programs you could join. It was very encouraged to be
- 29:01creative and part of that was like I, I remember now looking
- 29:05back, I can connect the dots a little.
- 29:07I started an entrepreneurship club at SAS because I was just
- 29:10so interested in being a founder.
- 29:12I started like a Junior Achievement chapter over there.
- 29:15We were, I mean, it was an interesting business.
- 29:18Like it was a pretty like brick and mortar type business that we
- 29:22built. It wasn't very and, and we were
- 29:25all very like mathy and sciencey people.
- 29:28So like we wanted to do something that was like
- 29:30chemistry and physics and I don't know what we were thinking
- 29:33in time. We were I I was a freshman and
- 29:35we ended up building this like take.
- 29:38We took Styrofoam and we were able to remold it and design it
- 29:42into like wasted Styrofoam, So like 0 cost and be able to
- 29:46remould it and design it into these like plastic coasters and
- 29:49stuff like that. And so that was like a business
- 29:51we ran for a while. We made like some like $500 or
- 29:54something, I don't know. Oh.
- 29:55OK. But yeah, it was like a nice way
- 29:58to get a group of people together and kind of build what
- 30:00a company was like and make money.
- 30:03And so that was like the beginning of it.
- 30:05And I think like slowly I found, as I found I, I, I had the
- 30:10courage to like, do my own thing as well.
- 30:13So sometime in like junior year, like I said, I watched that and
- 30:16I was like, oh, like I can build an app.
- 30:17And I think those small leaps of like, hey, you know, I can just
- 30:20do anything if I like set my mind to it.
- 30:22That kind of confidence came a lot from the school I was in.
- 30:26And seeing what the people around me were doing is like
- 30:29more of it was an internal confidence barrier of like, Oh
- 30:31my God, that seems so far away. But now I don't really tell
- 30:34myself that because I don't think anything is that far away.
- 30:37Like there's always a way to get there if someone else has so.
- 30:41I love that. I love that There's also a way
- 30:43to get there. OK, amazing.
- 30:44OK. And then last question before we
- 30:46wrap up here, what is your favorite Singaporean food that
- 30:51you miss the most now that you've moved back over here?
- 30:55I think Kaya toast. OK, OK, great.
- 31:01Great. Well, I think you can probably
- 31:02get some of the Kaya. Yeah, yeah, yeah, yeah.
- 31:05That's definitely something I haven't like, missed, missed in
- 31:07that way. But I yeah, Kaya toasted.
- 31:10I think as a kid I used to drink a lot of Milo.
- 31:12So I that's a very like, yes, you can get versions of it, but
- 31:16like, no, nothing tastes the same.
- 31:17Exactly. There's always a different type
- 31:19of chocolate powder and Milo's its own thing.
- 31:21Yeah. So.
- 31:22OK. Awesome.
- 31:23OK, Two of my favorite things as well.
- 31:25Nisha, thank you so much for joining us.
- 31:27Is there anything that you'd like to share with the audience
- 31:29before we close up here? No, not really.
- 31:32I mean, if there are any questions that you guys have or
- 31:35you find something interesting about Pepper, I'd love if you
- 31:37reach out. I'm on LinkedIn and I could
- 31:39leave my cow and leave in the comments.
- 31:42Awesome. Thank you so much, Nisha.
- 31:43Thanks for joining us and to our audience listening in, thanks
- 31:47for being with us this year. And if you like this episode,
- 31:52send us a like, drop us a comment and we will see you next
- 31:55time. Thank you so much.