67. From Vision to Reality: What it Takes to Build Successful AI Software | Nishit Asnani, Co-Founder of Sybill

Karthik Chidambaram: Hello, everyone. Welcome to a new episode of Driven by
DCKAP. I'm your host, Karthik Chidambaram, CEO of DCKAP. We make systems,
talk to each other, for manufacturers and distributors.

We are here in Mountain View, California. Being in Silicon Valley, let's
talk about AI. We are here with Nishit Asnani, Co-founder of Sibyll.ai.
Sibyll makes your life easier by listening to your sales calls, providing
summaries and actionable insights. On what you need to do with sales. Let's
talk about Sybill and more things.

Nishit thanks so much for joining this Driven podcast. Looking forward to
our conversation today.

Nishit Asnani: Yeah, I'm looking forward to our conversation today as well.
And I'm excited to join you, Karthik. Big fan of your work and super, super
awesome to be on this podcast. And I'm really happy to have you as a
customer as well.

Karthik Chidambaram: So, Nishit, thank you so much again. Thanks for your
time.

We are a happy customer with Sybill. Enjoy using your product. So thank you
for your great work. But tell us for the audience who are not very familiar
with Sybill, can you tell us about what Sybill does and how does it
differentiate from some other similar products in the market?

Nishit Asnani: Absolutely. So Sybill is essentially an AI assistant for
sales reps. The way we kind of think about it is it's like a Jarvis to the
Iron Man in you.

And the way it basically assists sales reps is by Being like an assistant
right next to you is there on every single call that you take. Every single
email that you send, it has visibility into your CRM, into your inbox. And
based on that, it helps you with a bunch of things. Firstly, it takes notes
during your calls.

It digests them into a format that is easily usable by everyone on your
team. sends updates to your manager, to your CRO, to your executives and so
on. It writes all of those updates right in your CRM, whether you use
Salesforce, HubSpot, whatever that is so that you everyone is up to date.
And in addition, it writes your follow up emails in your own style, in your
own tone of voice as well.

So essentially, it takes care of all the back end work so you can focus on
taking more customer calls and being more present in those conversations and
closing more deals.

Karthik Chidambaram: I love the actionable insights. You know, one way I use
the product is because obviously our teams do the calls and I really want to
see sometimes if you want to see what they really did instead of listening
through the whole call, I just listen to the I just read the simple summary
and it really helps.

And then I just go to that particular instant and use it. But is there a
reason why you guys chose sales when you were building this AI product?

Nishit Asnani: Yeah, absolutely. So I can go a little bit into the origin
story. Essentially essentially me and my co founders, we were actually
toying around with the idea of how we can make the digital world much more
accessible, especially during the pandemic.

So we started in the summer of 2020. And during that time, Gaurish, who was
my co founder and CEO, he was teaching a course at Stanford and after just
two classes of teaching online he was like it's so much easier to teach in
person because in virtual meetings even if all your students have their
cameras on you don't have the time or the Luxury to look at their faces and
figure out if your content is resonating or not And that gave birth to the
idea of that everyone's doing virtual meetings now, but there's a dearth of
tools that actually look at and understand what's happening in a
conversation What are the good bad and ugly parts of the conversation?

That's how the idea of Sybill took birth. What if we could understand
everything that's happening in a call, not just the transcription? But also
what does that what does everyone's intent look like? What does their body
language tell us about what they're going to do next? And so on. So we
started building on the system there.

And as we had conversations with people across the board from different
disciplines, different verticals, we realized that salespeople absolutely
love the idea of what we were trying to do. And they pounced on our initial
demos that hey, if you can build this out I can find a lot of value because
I want to understand what my prospect is thinking and doing during the call
and then after the call be able to communicate with them accordingly and
close the deals faster. So that was how we kind of got into B2B sales as our
target market and over the next couple of years Both Gaurish and I got on
like about 500 or so sales calls, with individual sales reps and their
managers and understood their entire breadth of problems and realized that
understanding.

People's behaviors and reactions in the call was a small part of the overall
puzzle, and we could really help them solve the entire gamut of things that
they have to do along with selling. So that's kind of like how we got into
B2B sales.

Karthik Chidambaram: So 2020 to 2022, you were doing more research, trying
to find the product market fit. What's really working, what niche area you
wanna get into, and that's when you decided you're gonna get into sales. Is
that right?

Nishit Asnani: Yeah. Yeah.

Karthik Chidambaram: And all the four co-founders were roommates.

Nishit Asnani: Yes. Yeah, yeah. We were all roommates and it's funny. So
Gaurish and I were classmates at Stanford, before we started Chamarco became
a roommate.

He was studying at UC San Diego and one mutual friend connected us to him
and then he eventually moved to the Bay Area. and started living with us and
also started working on the startup and Mehak who's our fourth co founder,
she is in fact Gorisha's sister and so she was doing her internship at
Harvard and then she moved here during the pandemic as well and we started
living together and then we started working together on what eventually
became Sybill.

Karthik Chidambaram: So Sybill.ai. Talking about AI, I can really see
there's a very practical use of how AI is being used, right? Because, you
know, so there's been a lot of buzz about AI, but where exactly are we using
it? This is a question a lot of people have, but essentially, right? So
Sybill is a great example of how it can practically make your life easier
where you listen through all the sales calls and you give us actionable
insights.

Hey, let's say if I need to send a note to the customer back, or if I need
to send a proposal, Okay. Sibyll tells me that, hey, I need to send a
proposal to the customer. And if everybody is connected in Sibyll, you know,
it just makes it more streamlined. So, what is the technology behind it,
right? So, when you guys decided, okay, two years you worked on the product,
and you said, okay, you're going to target the sales market.

So, what actually goes behind the scenes? Can you tell us about the
technology behind Sibyll?

Nishit Asnani: Absolutely. So the number one data source that we have is
your own customer calls. Right. And so in the customer calls, we're doing a
couple of things. We're understanding what each person is saying and
transcribing it at the same time.

Sibyll is also figuring out the levels of engagement and excitement that
each person is showing by looking at their smiling, nodding, and all of
these other facial cues. And then combining these two sources of
information, the non verbal and the verbal to figure out, okay, what is this
person's intent, how likely are they going to buy, what are the key things
that they are, they have pain points around, what are the key objections
that they had and so on.

And this and then, so we kind of encapsulate all of this data, figure out
what's most important and then use that to create a summary for the call.
That's how we get to the call outcome, the next steps, the pain points,
areas of interest and so on. Now what happens is once you have these
insights on each call Sybill also understands this data across the entire
deal.

So if you've had, let's say five calls and 15 email exchanges with let's say
Acme Corp, Sybill will understand all of that data in one go. And that based
on that, it will create a deal level summary for you which will correspond
to your own sales process. Whether you follow, let's say a band or a med
pick or any other sales process, Sybill can understand that and
contextualize the entire deal for you in that format. So you can take action
on it, whether you're the sales rep leading the deal or whether you're their
manager or whether you're the CEO who's looking at where each deal is at.

Karthik Chidambaram: So you talked about nonverbal cues. I mean, obviously
there's a lot of diversity involved there, you know, male, female, how are
you guys doing that?

You know, how are you analyzing the video? Is it your own software or are
you using a third party tool? How are you also making sure that diversity is
taken into account there?

Nishit Asnani: Absolutely. So it is primarily our own software that we've
built over the first couple of years. So, a lot of the time it's not about
going from like a raw video of a face to figure to engagement. There's a lot
of steps in between.

So the first thing is we try to understand the key points on the face and
there's like some 34 key points on the face that you need to kind of have an
understanding of and track for every single millisecond as to how the face
shape is changing and not just the face, but also the rest of the upper body
that that you show in zoom or g meet meetings.

Then from there as we track these points, then we have algorithms that, so
from these points and these kind of temporal movements, try to figure out
whether this person was nodding or smiling or frowning or leaning away or
leaning in and so on, and capture a bunch of these nonverbal reactions. And
then we have further algorithms on top that aggregate these nonverbal
reactions to then label every single moment in the call as, is this person
interested, is this person excited, or frustrated, or engaged, or
distracted, whatever that is.

And that is what you eventually see in the call summaries. We're able to
then contextualize these reactions to what was being discussed in the call.
So if, let's say, you were nodding heavily when I was demoing the HubSpot
integration to you then Sybill will be reasonable in trying to deduce that
you are somehow interested in the HubSpot integration.

And then it will find more evidence from the call to figure out if that is
true or not.

Karthik Chidambaram: And you talked about transcribing the calls as well,
right? Is that, again, your own technology or using somebody else's product
and then using APIs and doing that? How are you doing that?

Nishit Asnani: Yeah, so we have third party vendors that help us in
transcription, we have our own technology behind it as well to make sure
that all, like, the right sentences are attributed to the right speakers,
and then we have a further pipeline where we figure out the intent of each
person, we figure out the The sort of social hierarchy within the buying
committee if there's multiple buyers out there and to figure out the
strength of the buying intent and that eventually leads to the summaries

If you remember in the old days, let's say if I really want to get into tech
or software, they say you got to really be good at cc. But now AI is a new
thing.

Karthik Chidambaram: Let's say if I want to be really really good at AI.
What should I start with?

Nishit Asnani: Be really good at using AI first and foremost before building
it. I think there's a- You can cover a lot of ground just by getting really
good at prompting these LLMs like Chat GPT and Cloud in a way that you're
able to get them to do the stuff for you.

What I've found is like a lot of people try to be like, okay, how can I make
a cloud? But before that, can you get cloud or Chat GPT to do the stuff that
you're trying to do? An example recently was that I was trying to prepare
for a customer call and it was a major customer call and I was like, I know
this person can become a detractor if I don't take this well and answer
their objections properly.

So it's not like I need to fine tune an LLM to do anything of that sort. I
basically sat down and spent like 15-20 minutes just writing an amazing
prompt that gave ChatGPT all the context. about the person, about their
company, about my relationship with the person and the company about my
product, about how we sell them, about what I want out of the conversation,
and what are the major objections that I think we are going to face.

And I traded a couple of times on that and I was able to with chat gpt's
audio is able to run a pretty solid role play with it and role playing with
it for like 15 20 minutes gave it gave me the confidence that hey i can ace
this call and that's what happened as well so it's just first get really
good at using it get really good at figuring out what are the gaps here and
and then that can help you get better at programming over time because Now,
I don't think you should be starting from scratch.

You can actually learn a lot just by interacting with ChatGPT, asking it to
give you constructive feedback, being a tutor, being like an interactive
sort of teacher for you. And that way you can get really, really better at
programming.

Karthik Chidambaram: So prompting is very important. So you’ve got to really
start using AI first to build AI.

Nishit Asnani: Yeah.

Karthik Chidambaram: So you guys are doing this for sales. So let's say I
want to do this for HR or somebody else wants to do it for something else.
What should they do? How can they copy you?

Nishit Asnani: I think it's very important to figure out whether, like,
what's the simplest solution to whatever problem you want to solve at the
end of the day.

I think it's definitely about sales or HR, whatever vertical that is, but
it's way more about what's the simplest solution to the problems that I want
to solve for the customers. And if the solution is simply having smart
prompts for your LLM, then that's the easiest. Obviously that does not solve
all the major queries that your customers may have.

And so you may decide that, okay, I want to go deeper. So then beyond that,
then you have to build retrieval systems. So, what's basically a trend over
the last one and a half years as people are building retrieval, augmented
generation or rag systems across the board. And there are lots of flavors of
how you can build a rag system.

A rag system is basically. A system that digests all of the data context
documents that you have about about the space Or about let's say in sibyl
scenario about a specific sales deal and organizes it in a fashion That it's
easily accessible and then whenever the customer asks a question or wants to
fill out a certain thing like hey I'm going to What's the budget for this
deal?

Then the RAC system goes on, accesses all that information, and tries to
figure out what is the most relevant part of information that can help
answer this question, and then comes back with it. So it's different from
using an LLM right off the bat because you have to give it a lot more
context for the RAC to be effective.

So yeah, that's the next level of doing things. And then even further than
that, the next level is, if you really want to get into and fine tune the
model, then you'll have to create, compile your own data sets, fine tune the
LLM to be much more specific to your domain, and then serve the customers
based on that.

Karthik Chidambaram: So how long did it take for you guys to build the
initial product? Because you said for a couple of years you did market
research, you went and talked to customers, talked to a lot of salespeople,
tried to find it. Okay, you decided that. Sales is what I'm going to attack
and tackle. How long did it take for you to build the actual core product?

Nishit Asnani: Yeah, so we we actually started Building the product as soon
as we started the company And so while we were having a lot of conversations
with the customers We were also building the product and testing it out on
the site, even though we were not charging for it That's how we got some of
our early users It was primarily brute force just reaching out to people on
linkedin and being like hey, do you want to?

Try this out. There's something new and then as people used it We got to
learn a lot more about like what needs to be there in the core product as
we're talking about, right? Like we can't just be building a core product
with drops of engagement We need to go deeper and deeper and create more
interactive summaries and things like that.

So that's kind of like how we got on that road. And we already had a
summarization platform before the end of 2022, which is when chatgbd
launched. But like, so we were able to figure out the key components of a
call, the buying intent, the social hierarchy, the political discourse, and
so on from a conversation.

But the summaries were not the most easily readable. And then, and we had a
few early customers as well, and then the flip switched because once GPT 3.
5 was launched, which is the model that powers that power tragedy at that
point we saw that this could cause a step change and how digestible our call
summaries were.

So we plugged it at the end of our AI pipeline, got amazing results, and we
were the first in the sales tech space to launch call summaries back in
December of 2022. And, and then we quickly realized that we, let's say we
nailed call summaries. What do we do next? We after digesting a summary of a
call, the sales rep wants to send a follow up email, right?

They want to carry the conversation forward and move the deal forward. So we
started building follow up emails and then we started building deal
summaries and so on. So I would say that our core product, like there were
some inclination, inclinations of it in 2022, but a lot of the stuff got
built in 2023.

And now we are kind of. going to the next stage of what our core product is
going to look like. We've expanded the scope of what, what we, what we want
Cibyl to be doing based on what our customers have been telling us and where
the industry is going. And so, probably if you ask me the same question next
year, I would say the core product is being built right now in 2025.

Karthik Chidambaram: So you guys have done very impressive work in the last
four years. So let's say somebody is starting out. This is a problem a lot
of startups have. Hey, I built my product, but I'm not really getting
customers. So what do I do? You know, what am I doing wrong?

Nishit Asnani: So I think that's a whole that a lot of it, especially
technical founders do get into because they're much faster at building
product than getting customers. And we also had that problem for some time.
I think a lot of it has to do with like, as you're building the product, you
want to get users as soon as possible, even though your product looks ugly
and it may not solve all the pain points Your customer needs to get solved.

But still, you should keep hunting for users, get a few users, get their
feedback, and basically have them on as like a user or a customer advisory
board that you can ping every week, every two weeks, whenever you have
updates, and ask them for feedback. That's how you learn and eventually you
get to a point where a product is retentive and at least, there is at least
some non zero number of users who visit the product every single day or
every single week and and then there's a, you can iterate a lot more not
just based on qualitative feedback but actual usage data and based on how
users are interacting with it you can learn a lot more about what they want.

So I think it's a much more iterative process than being a step change that,
hey, I build a product first and then once the product is done then I try to
get customers. A lot of the time, especially in the early stages, the stage
where you don't have product market fit, it's much more iterative.

You build something, get some users, learn from it, learn what are the bad
parts of your baby, and then you fix them, and then you go back again and
again ask users for feedback.

Karthik Chidambaram: So keep hydrating.

Nishit Asnani: Yeah. Yeah.

Karthik Chidambaram: So you guys were doing about a hundred thousand dollars
in revenue around 2023. And from there you just got on to a million dollars.
So that's very, very fast growth. How did that happen?

Nishit Asnani: Yeah, I would say a lot of it is due to two factors. One is
that product innovation. We were the fastest to launch call summaries and we
were, and we're still the most accurate when it comes to the call summaries
that we have.

Similarly, we were the fastest to launch other defining AI products like AI
follow up emails and deal summaries as well. And we've continued to remain
the most innovative and on the frontier of sales tech since then. What that
means is that, and it's not just about being fastest in terms of developing
technology, but also integrating that into the, into the product in a way
that the product doesn't get too complex or hard to use.

It's still easy to use, very easy for a new person to onboard themselves and
get to the aha moment quickest. And that's what we focused on. And that's
helped us a lot in terms of growing that customer base quickly. Because our
promise is that, hey, you've got to use this amazing technology and this is
going to solve these very important use cases for you as a sales rep.

And then when a sales rep actually tries out the product, they see it happen
within the first couple of hours of using the product. So they want to keep
using it and keep spreading the word in their team and refer other people.
Which brings me to the second reason why we've grown fast, which is that our
customers are amazing.

And they have, they have not just Kind of adopted Sybill they've enjoyed
using the product and they've been vocal about how, how useful it's been for
them and they have and referrals has been a huge channel for us in order to
get new customers. Most of our customers have referred at least someone else
to try using the product.

Some of them have referred many, many, many people and that's helped us get
a lot of new revenue as well. And, like, while it's easy to say that, hey,
the product is great, I think a lot of the credit goes to our customers who
have been super supportive, super helpful. They've given us amazing feedback
and also great new customers as well through their efforts.

Karthik Chidambaram: So it's mostly product led growth so far, yeah?

Nishit Asnani: Yes.

Karthik Chidambaram: And how big is the technology team?

Nishit Asnani: So we are, I think, about a dozen folks in product and tech.
Yeah.

Karthik Chidambaram: Okay. Very interesting. Okay. Cool. So you guys also
raised, you guys recently raised 11 million dollars, so congratulations on
that.

What are you going to do with the money?

Nishit Asnani: So essentially if you think about our vision, we're trying to
be the Jarvis for a sales rep, which is, In the ideal world, we want to be a
thought partner. So if, let's say I'm a sales rep and I have several sitting
right next to me, I wanted to not just do the backend stuff for me, which is
just doing right now.

I also wanted to help me with deal strategies. I want it to help me with it.
This is the next step you should take to move this deal forward. I wanted to
also do a lot of the internal communication of like, When I want something
from product team, Sybill takes the initiative, goes in Slack, texts the
product team that hey, can you send me this or text the marketing team?

Hey, can you send this case study or can you make this case study and things
like that? So it's kind of like a true assistant that I can ask to get stuff
done. It also creates collaterals for me, creates pricing codes, creates
kind of competitive battle cards and so on. And we're not there yet.

We're probably like 5 percent of the way to becoming a true thought partner
or true AI assistant to the sales rep. And so the focus for this fundraise
is to build towards that. Get closer towards being a true AI assistant to
the sales rep. While also being super, super focused on our end user, which
is a sales rep.

And solving for them. We would consider ourselves really successful if we
are able to boost their their quota attainment and their income not just by
10, 20%, but hopefully by two to three X. That's kind of like what we want
to get towards. And that's a lofty goal that might take a few years, but we
believe we can get there.

And that's, that's our core focus with this race. So so for this, we would
need to obviously expand our customer base. We would need to build a lot of,
fundamental functionalities into the product and keep innovating on the
technology side of things.

Karthik Chidambaram: How do you allocate capital? You know, do you have,
like, a playbook in terms of, hey, this percentage is going to be put on
sales, this percentage is going to be put on product? How do you do that?

Nishit Asnani: Yeah that's a great question. So that's also something that
like, so it's a lot of it, like, okay, what are the key milestones that you
want to hit with this capital in the next couple of years? What would help
us get to the next stage and de risking the next set of big risks to the
company and then based on that Like what do we want to achieve?

On the product side We want to de risk building certain technologies and
building a product that's actually able to Have a certain win rate on the go
to market side We want to be able to build a scalable motion that helps us
get customers at a certain rate and limit churn and so it's about like
figuring out those milestones first and then going back from there as you
okay in order to hit them What are the Human as well as other resources that
we need to get there.

And obviously like our investors and advisors help a lot with that as well.

Karthik Chidambaram: So you live in the Bay Area and you operate from the
Bay Area. How important do you think is Bay Area today, right? Is it the
same as it was, let's say, 10 years ago where they say you have to be in the
Bay Area to run a product or, you know, it offers significant advantages. Do
you see the same thing today or?

Nishit Asnani: I think things have definitely changed from 10 years ago. In
the sense, in a few senses and in some other senses, they're basically the
same. So how have things changed? Is that I think especially in certain
domains, our talent is more distributed.

So earlier it used to be that you would only get really high quality
engineers in the Bay area. Now that's changed. Not necessarily the case. A
lot of the high quality engineers are still in the Bay Area, but you also
find more Talent elsewhere as well if you can look for it and and I think
Because of covid people have become more open To also buying more remotely
and that was basically the initial bet on which we started the company So in
terms of acquiring customers for sure It's much more, I would say,
democratized.

That being said, there are certain things that have stayed true even today.
There's, even today, there's a significant advantage, especially for a core
team to be in the Bay Area. Because like a lot of the industry events happen
here, a lot of the conversations with relevant veterans in the industry with
potential investors, with existing investors, a lot of those conversations
still happen here in person.

And you learn a lot just by interacting with people on a serendipitous note,
somewhere like walking down Castro street in Mountain View and you meet a
customer or you meet an advisor. And then you have a 10 minute conversation
and that opens up doors to new ways of thinking, and those are things that
were true 10 years ago.

They're still true today and that those I think are some significant
advantages that being said it's not it's not the end all be all and I think
great companies can be built from other places as well I definitely think
that being the Bay Area definitely helps us at Sybill.

Karthik Chidambaram: And you went to IIT Kanpur. Tell us about your IIT
Kanpur days. I have been to Kanpur as well. I went to Kalyanpuram. I went to
Kanpur. I also visited the Rajiv Motwani building. So it was a great
experience. Tell us about your experience at IIT Kanpur. I didn't study
there, but just went there. Tell us your experience of IIT Kanpur.

Nishit Asnani: Yeah, it was fantastic. I think the undergraduate years are
amazing because you develop as a person and figure out your own interests.
And for me, IIT Kanpur was an amazing experience for that, for those
reasons. So I was in the computer science department. I did my bachelor's in
CS and did a minor in English literature, which was also fun.

I'm very grateful that I had some amazing teachers, professors, and even
more grateful that I had an amazing group of friends. With whom I think
I've, I've learned a lot and that has shaped a lot of the a lot of what I am
today. I think some of the key, key takeaways for me of like, like why it
was amazing to be at IIT Kanpur is first and foremost, I think Just
developing that inner circle, that group of folks who are really curious,
really want to learn and and are intellectually driven and get stuff done
which was which is something that has shaped a part of my personality quite
a lot, and I wouldn't have been an entrepreneur if not for that.

Secondly, I think I learned a lot about myself and about what I enjoyed
doing. I went really deep into the underpinnings of computer science and
really enjoyed doing research in that field. I had some amazing mentors,
professors there that helped me get there. And I'll forever be grateful for
that because without that, I wouldn't have been here in the Bay Area and
would not have started Sibyl and truly was an amazing experience that way.

Kanpur itself is a very interesting city. It has its own quirks and its own
amazing adventures that you can have there. So, yeah.

Karthik Chidambaram: I love the experience. It also gets a little cold in
the night. I mean, it depends on the time of the year. Yeah, yeah, yeah.

Nishit Asnani: The winter can be brutal. Oh, yeah, yeah.

Karthik Chidambaram: So the theme of this podcast is driven. How are you
driven?

Nishit Asnani: I think I derive a lot of my drive from learning. So. I think
as long as I'm learning new things and exploring my curiosity, I feel like
I've accomplished something in the day and I'm, I'm being true to myself.

And so, and that's what I enjoy a lot about being a startup founder because
there's no shortage of learning. Everyday or every other day you get hit in
the face with a new almost company killing problem to solve and then you
have to solve it and then you solve it and then you move on. And that's I
think the accumulation of these experiences is what eventually helps you
truly get better Hopefully, hopefully we are doing better, from our mistakes
and and also add true value to the people who work with you As in like the
other folks the team members as well as the people who you're serving as
customers So I truly enjoy the learning and the impact.

And that's what drives me every day. The other thing I think is I am fairly
competitive and winning really drives me quite a bit as well. So closing a
deal with an important customer or closing a really important hire and these
kind of things also give me a lot of joy. And not just because I'm learning
a lot, but also it gives me a sense of accomplishment that, okay, we've,
we've, we move one step closer to de risking something important or to
hitting the next milestone.

Karthik Chidambaram: You write poetry as well. What do you do outside of
work? Poems?

Nishit Asnani: Sometimes, yes. In general, I enjoy reading and writing a
lot. So, back in undergrad, I used to write a lot of short stories. Even
before that, when I was in school, back in Bhopal, I used to write novels.
Over time, I think, I haven't had the time to write novels, but I've
definitely had the time to write poetry sometimes.

In my free time, like apart from writing and reading, I do enjoy going on
drives. California is a great place to do that. So yeah, enjoy being in
nature, close to the beach and so on and having a good time with friends and
family.

Karthik Chidambaram: Let's talk a little bit about writing, because I think
that's very interesting. Do you guys have a writing culture at Sybill and
how important is writing?

Nishit Asnani: I think it really varies across the team. Some, like for me,
writing is important in the work context, particularly because. I think I
think better when I write down my thoughts and writing really helps me
refine my thought process and get to the depths of what I really want to get
to, which just kind of brainstorming my head sometimes doesn't get me there.

And then secondly also, there was also one of the reasons why I started
leading marketing at Sybill because I really enjoyed writing and I thought I
could have a huge impact trying to attract a new set of customers to Sybill.
And I enjoy that part of my day to day work quite a lot. Copywriting is fun,
especially more so when you get results out of it.

So that's there. I think we do have a I wouldn't say we have like a very
strict writing culture at Sybill, but we do have a pretty decent culture
where people, because like half of our team is remote, we do have a pretty
strong culture where people are able to communicate properly on Slack and
email and so on and express their thoughts.

We do use asynchronous messaging. We do use asynchronous video messaging
quite a lot as well. Sharing Loom videos with teammates to kind of move
projects further. And I think that's pretty efficient. Yeah.

Karthik Chidambaram: Interesting. And Cool.

So let's stick to writing. So you write a journal as well?

Nishit Asnani: Yes.

Karthik Chidambaram: And how long have you been writing journals?

Nishit Asnani: I think it's been at least 13ish years.

Karthik Chidambaram: 13 ish years? Yeah. Oh, that's amazing. So you write
every day?

Nishit Asnani: I don't write every day. I definitely write when there's like
a big moment and I'm feeling emotionally overwhelmed. Really happy or really
sad moments. And I do write when I want to clear my head and put some
thoughts on paper.

Karthik Chidambaram: And you have the, let's say you want to go back like 10
years, you know, you can do that easily? You have all that?

Nishit Asnani: Yes.

Karthik Chidambaram: Very nice. So we work with a lot of manufacturers and
distributors. You guys do Sybill for sales. Can you tell us some practical
uses of how manufacturers and distributors can leverage AI for their day to
day work?

Nishit Asnani: Absolutely. I think there's a lot of low hanging fruit when
it comes to leveraging AI for day to day work. So for instance, if you were
taking a bunch of customer calls or partner calls on a daily basis,
obviously use Sybill to record the calls and get call summaries.

But even beyond that as I was telling earlier you can use with like an audio
interface to do role plays or coach yourself on how you can do better.
Something that I've done in the past that's helped me quite a lot is like
taking the transcript of a customer call that I had, or let's say a sales
rep on my team had, feeding it into An LLM like Chat GPT or Cloud and asking
it a very pointed very pointed set of questions To give me feedback on what
I could have done better or what kind of questions could I have asked that I
did not ask in my discovery conversations?

That helps me get better as a seller and also in general in terms of
communicating with potential customers so that those are like some very
baseline things some of the other things i've used it for obviously like a
lot of people say about content generation and the A lot of people try to
generate content for a website and blogs and LinkedIn and so on to grow
their presence I think These systems are getting better at it, but they're
still not there where I mean, you can still distinguish between authentic
content written by a Karthik.

If you were to write a LinkedIn post, that's still, you can see that it's
more authentic. But there are certain other types of content where AI is
super impactful. There are tools right now that you can use, like Superhuman
and Fixer, that help you respond to emails. And they are pretty on point and
follow your style of how you respond to emails.

You can obviously use Sybill for following up on customer conversations and
those emails, and so on. You can use these other tools for email follow ups.
And you can also use AI to generate day to day content, like if, if you're
someone who needs to generate project proposals every day, you can use AI to
generate those as long as you have a really strong prompt.

And what I see is a lot of people give up on AI for generating content or
for responding to emails or for doing having feedback and so on because they
give a first prompt and the answer they get from Chat GPT is pretty generic
and the key to unlocking the real value is spend more time in honing the
prompt and the context that you're giving the AI system instead of spending
a minute to write a prompt spend 15 minutes to write that give yourself that
target that helps you because If your end goal is that you want to be much
more efficient, there's no better hack right now than to use AI to solve a
bunch of your problems that you can.

And so it's worthwhile investing that much time and effort into creating the
right prompt, creating the right context, giving it feedback, helping it,
helping it help you.

Karthik Chidambaram: What sort of content generation? Do you see any other
use cases, let's say, at manufacturing or at distribution?

Nishit Asnani: So as I said, like content generation is one part of the
things.

Another thing is like, as we're trying to build civil towards a thought
partner, that's, that's important. Like, for instance I used, uh, chat GPT
just a couple of days back to go back and forth on my marketing strategy for
the next quarter. It was like, I had a few different ideas and a few
different problems that we faced in the previous quarter.

And I was like, hey, based on this can you can you help me come up with
okay? What should we even try doing and give it the constraints that hey,
these are the human constraints. These are the capital constraints. What can
we do here got some initial ideas at the end of the day? It's my job to
craft that strategy, but AI can help get me started a few weeks back I used
it to get some ideas on Create on like creative ways to leverage an event I
was going to an event and I was like, I just don't want to be another face
among 2, 000 people I want to Go there with something interesting of an
interesting value proposition an interesting piece of collateral What could
that be and it gave me a bunch of ideas.

So so so those are just these things in like everyday life that as a
Distributor like there are things that you there are problems that you need
to solve and with every single problem You should probably think about hey
If it is going to take me two or three hours, is there any way I can reduce
that time to half an hour or 15 minutes using AI?

And if that is possible, what would I need to get there? Another, I think
less used aspect of AI tools is that if you connect them to the right data
source, and this comes back to your value proposition as DC cap, right? If
you connect them to the right data sources, you are not just restricted to
the prompts that you give, You can actually use the data that you have in
your own systems of record and get amazing output out of it.

So for instance, like a very popular use case is if you connect your to your
customer support engine and to your CRM, let's say HubSpot and let's say to
your knowledge base and then ask it questions that at the intersection of
all of these like, hey, this customer is asking for this particular feature
and how to use it.

Do we have any documentation on it? And the AI system can help you answer
it. This can really speed up and make things easier. The job of your team or
your or your own self and your own conversations much more faster and
efficient.

Karthik Chidambaram: Make systems talk to each other.

Nishit Asnani: Exactly. Yeah, exactly.

Karthik Chidambaram: I would like to end with this question, Nishit. What
book are you reading right now?

Nishit Asnani: So I'm reading this book called The Innovator's Dilemma. I
think it's a very popular Silicon Valley textbook, so to say. But yeah, I'm
reading that because we're working on some company level strategy right now.
And I think it's a solid book written initially in the 90s. And so many of
the principles are still valid today.

Karthik Chidambaram: So Nishit, thank you so much for your time. Really
enjoyed this conversation on Sybill and AI.. Thank you so much. Great
chatting with you.

Nishit Asnani: Great chatting with you as well, Karthik. Take care.