October 7, 2025 • 21 mins
In Episode 99, I do something a little different: I take you behind the scenes as I use ChatGPT in voice mode to design a real automation from start to finish. The goal? Build an internal chatbot for product and engineering that’s trained on CX call transcripts stored in Gong, so teams can ask targeted questions (“What’s frustrating customers in Module X?”) and get instant, concise answers with deep links back to the exact call moments.
You’ll hear how I frame the problem, push the model to avoid hallucinations, and pick a stack that balances speed, privacy, and scale: Gong → Airtable as the searchable store → a Zapier-hosted chatbot for querying. We also cover transcript hygiene (auto-removing small talk and personal details), vendor privacy considerations, and a simple habit hack: having AI remind you later to actually implement the ideas you generated while walking the dog.
I’ll link the step-by-step PDF I asked ChatGPT to generate in the show notes so you can follow along and adapt it to your environment.
If this sparks ideas for your own digital CX programs, follow/subscribe and drop a review—it really helps more practitioners find the show.
+++++++++++++++++
Like/Subscribe/Review:
If you are getting value from the show, please follow/subscribe so that you don't miss an episode and consider leaving us a review.
Website:
For more information about the show or to get in touch, visit DigitalCustomerSuccess.com.
Buy Alex a Cup of Coffee:
This show runs exclusively on caffeine - and lots of it. If you like what we're, consider supporting our habit by buying us a cup of coffee: https://bmc.link/dcsp
Thank you for all of your support!
The Digital Customer Success Podcast is hosted by Alex Turkovic
🎬 This content was edited by Lifetime Value Media.
Learn more at: https://www.lifetimevaluemedia.com
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
SPEAKER_01 (00:02):
Today we're going to use ChatGBT in voice mode to
help us automate.
Once again, welcome to theDigital Customer Experience
podcast with me, Alex Turkovich.
So glad you could join us heretoday and every week as we
explore how digital can helpenhance the customer and
employee experience.
My goal is to share what myguests and I have learned over
(00:25):
the years so that you can getthe insights that you need to
evolve your own digitalprograms.
If you'd like more info, need toget in touch, or sign up for the
weekly companion newsletter thathas additional articles and
resources in it, go todigitalcustomer success.com for
now.
Let's get started.
Hello and welcome.
My name is Alex Turkovich, andthis is the Digital CX Podcast,
(00:46):
the only show where we talkabout digital in CX on a weekly
basis, roughly.
Um welcome back to this isepisode 99.
Super exciting because, well,the next one's 100, which some
would argue is an arbitrarynumber, but uh still we gotta
celebrate our milestones alittle bit here and there.
Um, today I have a pretty coolepisode for you that is kind of
(01:08):
inspired by one of my manyworkflows.
Um, specifically with regards tovoice mode in Chat GPT.
Sometimes when I'm on a walk orjust doing other things, I like
to put Chat GPT in voice mode tohelp me get some things
accomplished.
(01:28):
And one of the things I've beendoing lately is having it help
me work through um partially inmy brain, but also get advice on
how to automate certain things.
And um, I was doing this theother day on a walk.
I was talking, I was chattingwith ChatGPT about how to
automate a certain process forthis podcast, actually.
(01:51):
And it dawned on me that um someof you may not be super familiar
with that mode.
And if you are, awesome.
If you aren't, welcome to theworld um of talking with AI in
real time.
But what I figured I would do isactually have one of these chats
live on the podcast uh with youlistening in.
(02:12):
So the scenario that I'm puttingforth uh forward with ChatGPT is
I want it to help me create aautomation that pulls chat
transcripts or or voice calltranscripts, customer call
transcripts into a database sothat our product team can query
it for product feedback.
Um which is actually a prettycool use case, right?
(02:32):
You you could theoretically Imean you can set up a a chat bot
for product teams to use whenthey want feet like kind of
quote unquote live customerfeedback on specific modules and
things that are not working wellwith them, things that are
working, um, and those kinds ofthings, right?
So um you'll hear that wholescenario that I lay out with the
(02:54):
chat.
A couple things to point out.
Do listen for how I'm promptingit, especially in the beginning.
I'm giving it very specificinstructions for what expertise
I want it to have.
I'm also explicitly telling itnot to make stuff up and do some
research on the tools and thecapabilities.
Um, a lot of times when I dothis, I go uh a little bit more
(03:18):
into detail, but I didn't Ididn't want to present you with
like a 20 or 30 minute longchat, right?
I wanted to keep it under 15minutes for you.
Um but sometimes I will gofurther into detail about you
know the specifics of setting acertain part of the automation
up.
Um and in this case, what Iasked it to do at the end was to
(03:41):
spit out a PDF kind of guide forhow to set that up.
And I will link that PDF in theshow notes if you do want to um
look at it afterwards and seeactually what it came up with.
Now, I uh you know, sometimesyou have to take those things
with a little bit of a grain ofsalt.
Uh you know, it it it won't behyper accurate, especially if
(04:05):
you haven't given Chat GPT uh orClaude or Gemini for that matter
the specifics of what it isyou're doing, what environment
you're in, what your role is,who you are, who the company is.
You know, but if you have spentenough time with any of those
models or tools, they'll have ageneral context for who you are
(04:26):
and what you do.
And so the the results will be alittle bit more tailored towards
you.
Um now I think we all know whenwe ask, you know, Chat GPT,
especially to pump out like aPDF or whatever, it's gonna give
you the bare bones, right?
It's not gonna give you anythinghighly formatted or whatever,
but um I think Claude especiallyis making a lot of inroads into
(04:46):
making that a little bit better,especially with regards to
PowerPoint generation and thosekinds of things.
Anyway, I'm rambling.
What follows is my roughly 10,12 minute um conversation that I
just had with Chat GPT aboutsetting this thing up.
I hope it's useful to you, and Iwould love your feedback on
maybe some things I could havedone differently or different
(05:08):
ways of, you know, prompting itto get a better result and those
kinds of things.
I'm always open to feedback,right?
So if you guys are are workingon this kind of stuff and doing
this on the daily basis, youknow, we can all learn from each
other in terms of betterprompting.
But I hope that's insightful andI hope that you can apply this
to your own role, right?
The end goal is for you to alsothen do this kind of stuff so
(05:29):
that as you're doing the dishesor as you're taking the dog for
a walk or as you're driving homefrom work, whatever it is you
do, you might as well be alittle productive at the same
time.
Now, the trick is after you havethese conversations, going back
and actually executing on them,right?
I've had a lot of these types ofconversations with Chat GPT, and
(05:51):
then after I get home orwhatever, I just kind of forget
about it and I don't go back toit, right?
So whatever you do, make sureyou make give yourself a mental
note to go after it.
And in fact, one thing that youcan do at the end of these, um
at the end of these chats is youcan have ChatGPT send you a
(06:12):
reminder at I don't know, 8 p.m.
that evening or you know ninea.m.
the next morning that you needto action on what it is you
chatted about, right?
So you can actually have itremind you.
Um especially if you're workingwith calendar integrations and
whatnot.
But anyway, here's theconversation.
I hope you enjoy it.
(06:34):
Okay, so I'm hoping you can helpme with a little project,
ChatGPT, where I need you toessentially act as a digital
automations expert, specificallyAI automations expert with a
deep knowledge in customerexperience best practices.
SPEAKER_00 (06:52):
Absolutely.
I can put on my AI automationsexpert hat for this.
Just let me know the details ofwhat you're working on and what
kind of Bidens or setup youneed.
We can dive right in from there.
SPEAKER_01 (07:03):
Okay, the other thing I want you to do is to
make sure that as we'rediscussing these things, you're
conducting some deep research onbest practices and tooling
capabilities, specifically withregards to API availability.
And then also don't make stuffup.
So I want your answers to begrounded in nothing but reality.
SPEAKER_00 (07:27):
Absolutely.
I'll keep everything as real andgrounded as possible.
So as we go along, I'll do thedeep research on those best
practices and the toolingdetails, especially around APIs.
And of course, I'll stickstrictly to what's actually
true.
No fictional stuff, justreality.
Let's get into it.
SPEAKER_01 (07:43):
Okay, sounds good.
So my goal is to essentiallycreate an internal chatbot for
internal use only, primarily bythe product and engineering
group.
Um, and this chatbot should betrained on customer experience
team call transcripts for theintent of giving product teams a
(08:06):
way to get instantaneous productfeedback across all of the
calls.
So the idea would be that aproduct manager could
essentially go to this chat bot,ask a question about maybe a
specific product or specificmodule, maybe a specific
frustration, or things thatcustomers are challenged with or
(08:27):
things that customers are reallyexceeding at and be able to get
instantaneous feedback based onthose call transcripts, but then
also to get links to thespecific calls where those um
where those things took place.
Now, in order to do this, um, Ican give you what our tech stack
(08:52):
is like, but then I would alsolike for you to ask me
clarifying questions one by oneuntil you have 95% um until you
have 90 95% clarity on what myend goal is and what we need to
build.
SPEAKER_00 (09:08):
Absolutely.
That all makes perfect sense.
And I love the approach of doingthis step by step with
clarifying questions.
So let's dive in.
First, a quick clarifyingquestion, just so I have a solid
foundation.
Can you let me know what yourcurrent tech stack is,
especially the platforms ortools you're using for storing
these call transcripts?
For example, are they housed ina system like Gong or in a CRM
(09:28):
or just in a database?
Once I know that, we can figureout the best ways to integrate
and make that chatbot a reality.
SPEAKER_01 (09:34):
Call transcripts, all inbound and outbound, are
stored in Gong.
Um and my initial intent was toperhaps create some sort of
export automation into somethingeither like Notion or Airtable
or Google Sheets that thischatbot can then reference.
(09:57):
And I would love your feedbackas we talk on if you think
that's the right approach.
SPEAKER_00 (10:03):
Perfect.
That gives me a great startingpoint.
So we know we're working withGong for the call transcripts,
and we want to create anautomation that exports that
data into a platform likeNotion, Airtable, or Google
Sheets that the chatbot canreference.
So here's my next clarifyingquestion to get us closer to
that 95% clarity.
Among those options, Notion,Airtable, Google Sheets, do you
have a preferred tool that yourteam is already comfortable
(10:25):
with?
Or are you looking for arecommendation on which one
might be easiest for the chatbotto query and for you to
maintain?
Let me know your preference andwe'll keep going from there.
SPEAKER_01 (10:34):
Looking for a little bit of a recommendation, but I
think from what I know, Airtableis probably the best suited for
that, just because it does playwell with integration tools like
make.com or Zapier.
And it also is a lot morescalable when it comes to
feeding it tons of data and andmaking it easily organizable.
(10:55):
Um that said I'm open to Notionor Google Sheets as well, if
those are things you'd recommendover using Airtable.
SPEAKER_00 (11:05):
Absolutely.
Airtable is a great choice.
It's really flexible.
And like you said, it playsnicely with integration tools
like make.com or Zapier.
It's also easy to search andorganize, which is going to be
really helpful for a chatbotthat needs to pull in customer
feedback details.
So I think that's a solid route.
Now that we have that piececlarified, let me ask another
question to get us closer tothat full clarity.
(11:26):
Do you have a preference for howthe chatbot itself is built or
hosted?
For example, are you thinking ofusing a specific chatbot
framework or platform?
Or are you open torecommendations on that as well?
Let me know.
We'll keep refining.
SPEAKER_01 (11:38):
I'm very open to recommendations on the chatbot
interface, whether it'ssomething we built from scratch
or whether it's an existing toolthat we can use.
Um preferably something that iseasily integratable with Zapier,
for instance, or make.com.
SPEAKER_00 (11:58):
Absolutely.
That helps a lot.
If we want something that'seasily integratable and works
well with Zapier or Make.com,one strong recommendation would
be looking at something like ano-code or low-code chatbot
platform.
A tool like LANBot or even amore flexible platform like bot
press can be really good becausethey let you integrate with
external data sources andconnect with webhooks or APIs to
(12:19):
something like Airtable.
So I'd recommend using aplatform like LANBot for the
chatbot interface.
It's pretty user-friendly and itcan call out to Airtable through
Zapier or make.com to fetchthose call transcript details.
That way, when a product managerasks the chatbot something, it
can just pull the relevant inforight out of Airtable and even
link them back to the originalgun call if needed.
(12:39):
Let me know if that sounds good.
And if it does, we can talk nextto us.
SPEAKER_01 (12:42):
I have a question for you about data privacy.
If we are to use one of theseexternal chatbot tools, are
there data privacy concerns thatwe need to be careful of in
terms of vetting that with thisvendor?
And if there are, are thereoptions for maybe native chatbot
capabilities within like Zapieror make.com, for instance?
SPEAKER_00 (13:05):
Absolutely.
And that's a really smartquestion to ask.
So in terms of data privacy,whenever you're using an
external tool, especially onethat's handling potentially
sensitive customer data fromcall transcripts, you definitely
want to vet that vendor fortheir data handling policies.
Look for things like GDPRcompliance, CCPA compliance if
you're in California, andgenerally how they store and
(13:26):
secure the data.
So that's something you don'thave to do a little due
diligence on.
Now, if you're looking for amore privacy-controlled option,
there are indeed some nativecapabilities.
For example, you can actuallybuild a simple chatbot directly
using Zapier's interfaces, oryou could use make.com with a
webhook and a custom front endthat you control.
That way all the data stays abit more within your own
ecosystem and you have a littlebit more control over privacy.
(13:48):
It might take a tiny bit moresetup, but it gives you that
extra layer of assurance.
So overall, you can definitelydo it with a vendor tool as long
as you vet them.
Or you can go the more nativeroute with Zapier make.
Let me know which one feels likea better fit, and we'll keep
going from there.
SPEAKER_01 (14:01):
Okay.
Since we are already a Zapiercustomer, we'd probably prefer
to build that chatbot interfacewithin Zapier.
Um and so the flow now, as Iunderstand it, would be that we
would feed the call transcriptsfrom Gong into Airtable and then
use Airtable to train thechatbot.
(14:23):
Can you confirm that?
And are there any other thingsthat you would need clarity on
before building out a wholeinfrastructure document for
this?
SPEAKER_00 (14:34):
Yes, absolutely.
And I can confirm that that flowsounds exactly right.
So essentially you're going tobe pulling those call
transcripts from Bung intoAirtable.
Airtable is going to be yourdatabase that the Zapier based
chatbot will query.
And then Zapier will act as thelogic layer to let the chatbot
retrieve the right informationand show it to your product
team.
So that's a solid plan.
I think we're pretty close tothat 95% clarity.
(14:55):
The only other thing I mightwant to know, just to finalize
everything, is if you have anyspecific requirements for how
the chatbot should respond.
For example, do you want it togive a summary of customer
sentiment or just list out thecall or offer a combination of
both?
Once we know that, we can puttogether a nice infrastructure
document and you'll be good togo.
SPEAKER_01 (15:13):
I'd like the chatbot to respond pretty directly with
specific feedback based on thequery from the product team,
which also means that I wouldideally want to strip some of
the call transcripts of any kindof small talk or personal
details and things like that aswell.
So I would like your input onhow we could potentially do that
(15:35):
in an automated fashion becauseit's unrealistic for us to go
and scrape hundreds and hundredsof call transcripts.
SPEAKER_00 (16:17):
Absolutely.
That all makes perfect sense.
And I love the focus on keepingit concise, but also allowing
deep links for more context.
So let's break that down intothe final pieces.
First, on the automatedstripping of small talk or
personal details.
One approach is to use a naturallanguage processing step in your
automation.
For example, you could have astep in Zapier that runs the
transcript through an AI textcleaner.
(16:39):
There are some pre-built toolsor even just a simple custom GPT
model that can identify andremove filler conversation.
That way you're only keeping thecore feedback.
We can definitely outline thatin the infrastructure doc.
Second, on the output side, it'stotally doable to have the
chatbot give a concise summaryand then provide a link directly
to the relevant call in Gong.
That way, product managers canclick through if they want to
(17:00):
hear the full context.
We'll just set up the chatbotresponses so they have a short
summary plus a listen to thefull call link.
With those final pieces inplace, I think we have
everything we need.
I'll go ahead and draft out aninfrastructure document that
covers the entire flow, pullingdata from Gong, cleaning it up,
storing it in Airtable, and thenusing Zapier to power the
chatbot interface.
If that all sounds good, we'llget that ready for you.
SPEAKER_01 (17:21):
That sounds good.
When you create this document,um ideally please um provide it
in PDF format.
And I would like for you to beas specific as possible when it
comes to the setup of each step.
So for example, in Airtable, Iwould want you to be very
specific in terms of whichcolumns we need to set up for
the data being pulled.
(17:41):
For instance, a link to, youknow, um the call transcript
itself and those kinds ofthings.
So be very specific in thosesteps and essentially treat it
as if you're teaching a beginnerhow to set up this kind of
automation with each tool.
SPEAKER_00 (17:57):
Absolutely.
I'll make it asbeginner-friendly and detailed
as possible.
I'll include all the specificsfor each step, like how to set
up your air table columns, whereto link the call transcripts,
and how to configure each partof the Zapier flow so that it's
all crystal clear and a nice PDFfor you.
So I'll put that together andwe'll have you all set up big
go.
SPEAKER_01 (18:14):
Great.
Thank you.
SPEAKER_00 (18:16):
You're very welcome.
I'm glad we could get everythingironed out.
I'll get that detailed PDF readyfor you with all the
step-by-step instructions.
And if you have any morequestions along the way, just
let me know.
Happy to help.
SPEAKER_01 (18:29):
Okay, so that is essentially the conversation I
had with it.
Again, I will put the outputs,that PDF output of what it came
back with in the show notesbelow so that if you want to go
look at it, you definitely can.
Uh again, would love yourfeedback and would love to hear
from you what kinds of thingsyou're chatting with ChatGPT
(18:50):
about on a regular basis.
But that'll do it for thisepisode.
I want to thank you for joiningme today and uh hope your week
ahead is an awesome one.
We'll see you next week.
Thank you for joining me forthis episode of the Digital CX
Podcast.
If you like what we're doing, uhconsider leaving us a review on
your podcast platform of choice.
If you're watching on YouTube,leave a comment down below.
(19:12):
It really helps us to grow andprovide value to a broader
audience and get moreinformation about the show and
some of the other things thatwe're doing at digitalcustomer
success.com.
I'm Alex Turkovich.
Thanks so much for listening,and we'll talk to you next week.
Today we're going to use ChatGBT in voice mode to
help us automate.
Once again, welcome to theDigital Customer Experience
podcast with me, Alex Turkovich.
So glad you could join us heretoday and every week as we
explore how digital can helpenhance the customer and
employee experience.
My goal is to share what myguests and I have learned over
(00:25):
the years so that you can getthe insights that you need to
evolve your own digitalprograms.
If you'd like more info, need toget in touch, or sign up for the
weekly companion newsletter thathas additional articles and
resources in it, go todigitalcustomer success.com for
now.
Let's get started.
Hello and welcome.
My name is Alex Turkovich, andthis is the Digital CX Podcast,
(00:46):
the only show where we talkabout digital in CX on a weekly
basis, roughly.
Um welcome back to this isepisode 99.
Super exciting because, well,the next one's 100, which some
would argue is an arbitrarynumber, but uh still we gotta
celebrate our milestones alittle bit here and there.
Um, today I have a pretty coolepisode for you that is kind of
(01:08):
inspired by one of my manyworkflows.
Um, specifically with regards tovoice mode in Chat GPT.
Sometimes when I'm on a walk orjust doing other things, I like
to put Chat GPT in voice mode tohelp me get some things
accomplished.
(01:28):
And one of the things I've beendoing lately is having it help
me work through um partially inmy brain, but also get advice on
how to automate certain things.
And um, I was doing this theother day on a walk.
I was talking, I was chattingwith ChatGPT about how to
automate a certain process forthis podcast, actually.
(01:51):
And it dawned on me that um someof you may not be super familiar
with that mode.
And if you are, awesome.
If you aren't, welcome to theworld um of talking with AI in
real time.
But what I figured I would do isactually have one of these chats
live on the podcast uh with youlistening in.
(02:12):
So the scenario that I'm puttingforth uh forward with ChatGPT is
I want it to help me create aautomation that pulls chat
transcripts or or voice calltranscripts, customer call
transcripts into a database sothat our product team can query
it for product feedback.
Um which is actually a prettycool use case, right?
(02:32):
You you could theoretically Imean you can set up a a chat bot
for product teams to use whenthey want feet like kind of
quote unquote live customerfeedback on specific modules and
things that are not working wellwith them, things that are
working, um, and those kinds ofthings, right?
So um you'll hear that wholescenario that I lay out with the
(02:54):
chat.
A couple things to point out.
Do listen for how I'm promptingit, especially in the beginning.
I'm giving it very specificinstructions for what expertise
I want it to have.
I'm also explicitly telling itnot to make stuff up and do some
research on the tools and thecapabilities.
Um, a lot of times when I dothis, I go uh a little bit more
(03:18):
into detail, but I didn't Ididn't want to present you with
like a 20 or 30 minute longchat, right?
I wanted to keep it under 15minutes for you.
Um but sometimes I will gofurther into detail about you
know the specifics of setting acertain part of the automation
up.
Um and in this case, what Iasked it to do at the end was to
(03:41):
spit out a PDF kind of guide forhow to set that up.
And I will link that PDF in theshow notes if you do want to um
look at it afterwards and seeactually what it came up with.
Now, I uh you know, sometimesyou have to take those things
with a little bit of a grain ofsalt.
Uh you know, it it it won't behyper accurate, especially if
(04:05):
you haven't given Chat GPT uh orClaude or Gemini for that matter
the specifics of what it isyou're doing, what environment
you're in, what your role is,who you are, who the company is.
You know, but if you have spentenough time with any of those
models or tools, they'll have ageneral context for who you are
(04:26):
and what you do.
And so the the results will be alittle bit more tailored towards
you.
Um now I think we all know whenwe ask, you know, Chat GPT,
especially to pump out like aPDF or whatever, it's gonna give
you the bare bones, right?
It's not gonna give you anythinghighly formatted or whatever,
but um I think Claude especiallyis making a lot of inroads into
(04:46):
making that a little bit better,especially with regards to
PowerPoint generation and thosekinds of things.
Anyway, I'm rambling.
What follows is my roughly 10,12 minute um conversation that I
just had with Chat GPT aboutsetting this thing up.
I hope it's useful to you, and Iwould love your feedback on
maybe some things I could havedone differently or different
(05:08):
ways of, you know, prompting itto get a better result and those
kinds of things.
I'm always open to feedback,right?
So if you guys are are workingon this kind of stuff and doing
this on the daily basis, youknow, we can all learn from each
other in terms of betterprompting.
But I hope that's insightful andI hope that you can apply this
to your own role, right?
The end goal is for you to alsothen do this kind of stuff so
(05:29):
that as you're doing the dishesor as you're taking the dog for
a walk or as you're driving homefrom work, whatever it is you
do, you might as well be alittle productive at the same
time.
Now, the trick is after you havethese conversations, going back
and actually executing on them,right?
I've had a lot of these types ofconversations with Chat GPT, and
(05:51):
then after I get home orwhatever, I just kind of forget
about it and I don't go back toit, right?
So whatever you do, make sureyou make give yourself a mental
note to go after it.
And in fact, one thing that youcan do at the end of these, um
at the end of these chats is youcan have ChatGPT send you a
(06:12):
reminder at I don't know, 8 p.m.
that evening or you know ninea.m.
the next morning that you needto action on what it is you
chatted about, right?
So you can actually have itremind you.
Um especially if you're workingwith calendar integrations and
whatnot.
But anyway, here's theconversation.
I hope you enjoy it.
(06:34):
Okay, so I'm hoping you can helpme with a little project,
ChatGPT, where I need you toessentially act as a digital
automations expert, specificallyAI automations expert with a
deep knowledge in customerexperience best practices.
SPEAKER_00 (06:52):
Absolutely.
I can put on my AI automationsexpert hat for this.
Just let me know the details ofwhat you're working on and what
kind of Bidens or setup youneed.
We can dive right in from there.
SPEAKER_01 (07:03):
Okay, the other thing I want you to do is to
make sure that as we'rediscussing these things, you're
conducting some deep research onbest practices and tooling
capabilities, specifically withregards to API availability.
And then also don't make stuffup.
So I want your answers to begrounded in nothing but reality.
SPEAKER_00 (07:27):
Absolutely.
I'll keep everything as real andgrounded as possible.
So as we go along, I'll do thedeep research on those best
practices and the toolingdetails, especially around APIs.
And of course, I'll stickstrictly to what's actually
true.
No fictional stuff, justreality.
Let's get into it.
SPEAKER_01 (07:43):
Okay, sounds good.
So my goal is to essentiallycreate an internal chatbot for
internal use only, primarily bythe product and engineering
group.
Um, and this chatbot should betrained on customer experience
team call transcripts for theintent of giving product teams a
(08:06):
way to get instantaneous productfeedback across all of the
calls.
So the idea would be that aproduct manager could
essentially go to this chat bot,ask a question about maybe a
specific product or specificmodule, maybe a specific
frustration, or things thatcustomers are challenged with or
(08:27):
things that customers are reallyexceeding at and be able to get
instantaneous feedback based onthose call transcripts, but then
also to get links to thespecific calls where those um
where those things took place.
Now, in order to do this, um, Ican give you what our tech stack
(08:52):
is like, but then I would alsolike for you to ask me
clarifying questions one by oneuntil you have 95% um until you
have 90 95% clarity on what myend goal is and what we need to
build.
SPEAKER_00 (09:08):
Absolutely.
That all makes perfect sense.
And I love the approach of doingthis step by step with
clarifying questions.
So let's dive in.
First, a quick clarifyingquestion, just so I have a solid
foundation.
Can you let me know what yourcurrent tech stack is,
especially the platforms ortools you're using for storing
these call transcripts?
For example, are they housed ina system like Gong or in a CRM
(09:28):
or just in a database?
Once I know that, we can figureout the best ways to integrate
and make that chatbot a reality.
SPEAKER_01 (09:34):
Call transcripts, all inbound and outbound, are
stored in Gong.
Um and my initial intent was toperhaps create some sort of
export automation into somethingeither like Notion or Airtable
or Google Sheets that thischatbot can then reference.
(09:57):
And I would love your feedbackas we talk on if you think
that's the right approach.
SPEAKER_00 (10:03):
Perfect.
That gives me a great startingpoint.
So we know we're working withGong for the call transcripts,
and we want to create anautomation that exports that
data into a platform likeNotion, Airtable, or Google
Sheets that the chatbot canreference.
So here's my next clarifyingquestion to get us closer to
that 95% clarity.
Among those options, Notion,Airtable, Google Sheets, do you
have a preferred tool that yourteam is already comfortable
(10:25):
with?
Or are you looking for arecommendation on which one
might be easiest for the chatbotto query and for you to
maintain?
Let me know your preference andwe'll keep going from there.
SPEAKER_01 (10:34):
Looking for a little bit of a recommendation, but I
think from what I know, Airtableis probably the best suited for
that, just because it does playwell with integration tools like
make.com or Zapier.
And it also is a lot morescalable when it comes to
feeding it tons of data and andmaking it easily organizable.
(10:55):
Um that said I'm open to Notionor Google Sheets as well, if
those are things you'd recommendover using Airtable.
SPEAKER_00 (11:05):
Absolutely.
Airtable is a great choice.
It's really flexible.
And like you said, it playsnicely with integration tools
like make.com or Zapier.
It's also easy to search andorganize, which is going to be
really helpful for a chatbotthat needs to pull in customer
feedback details.
So I think that's a solid route.
Now that we have that piececlarified, let me ask another
question to get us closer tothat full clarity.
(11:26):
Do you have a preference for howthe chatbot itself is built or
hosted?
For example, are you thinking ofusing a specific chatbot
framework or platform?
Or are you open torecommendations on that as well?
Let me know.
We'll keep refining.
SPEAKER_01 (11:38):
I'm very open to recommendations on the chatbot
interface, whether it'ssomething we built from scratch
or whether it's an existing toolthat we can use.
Um preferably something that iseasily integratable with Zapier,
for instance, or make.com.
SPEAKER_00 (11:58):
Absolutely.
That helps a lot.
If we want something that'seasily integratable and works
well with Zapier or Make.com,one strong recommendation would
be looking at something like ano-code or low-code chatbot
platform.
A tool like LANBot or even amore flexible platform like bot
press can be really good becausethey let you integrate with
external data sources andconnect with webhooks or APIs to
(12:19):
something like Airtable.
So I'd recommend using aplatform like LANBot for the
chatbot interface.
It's pretty user-friendly and itcan call out to Airtable through
Zapier or make.com to fetchthose call transcript details.
That way, when a product managerasks the chatbot something, it
can just pull the relevant inforight out of Airtable and even
link them back to the originalgun call if needed.
(12:39):
Let me know if that sounds good.
And if it does, we can talk nextto us.
SPEAKER_01 (12:42):
I have a question for you about data privacy.
If we are to use one of theseexternal chatbot tools, are
there data privacy concerns thatwe need to be careful of in
terms of vetting that with thisvendor?
And if there are, are thereoptions for maybe native chatbot
capabilities within like Zapieror make.com, for instance?
SPEAKER_00 (13:05):
Absolutely.
And that's a really smartquestion to ask.
So in terms of data privacy,whenever you're using an
external tool, especially onethat's handling potentially
sensitive customer data fromcall transcripts, you definitely
want to vet that vendor fortheir data handling policies.
Look for things like GDPRcompliance, CCPA compliance if
you're in California, andgenerally how they store and
(13:26):
secure the data.
So that's something you don'thave to do a little due
diligence on.
Now, if you're looking for amore privacy-controlled option,
there are indeed some nativecapabilities.
For example, you can actuallybuild a simple chatbot directly
using Zapier's interfaces, oryou could use make.com with a
webhook and a custom front endthat you control.
That way all the data stays abit more within your own
ecosystem and you have a littlebit more control over privacy.
(13:48):
It might take a tiny bit moresetup, but it gives you that
extra layer of assurance.
So overall, you can definitelydo it with a vendor tool as long
as you vet them.
Or you can go the more nativeroute with Zapier make.
Let me know which one feels likea better fit, and we'll keep
going from there.
SPEAKER_01 (14:01):
Okay.
Since we are already a Zapiercustomer, we'd probably prefer
to build that chatbot interfacewithin Zapier.
Um and so the flow now, as Iunderstand it, would be that we
would feed the call transcriptsfrom Gong into Airtable and then
use Airtable to train thechatbot.
(14:23):
Can you confirm that?
And are there any other thingsthat you would need clarity on
before building out a wholeinfrastructure document for
this?
SPEAKER_00 (14:34):
Yes, absolutely.
And I can confirm that that flowsounds exactly right.
So essentially you're going tobe pulling those call
transcripts from Bung intoAirtable.
Airtable is going to be yourdatabase that the Zapier based
chatbot will query.
And then Zapier will act as thelogic layer to let the chatbot
retrieve the right informationand show it to your product
team.
So that's a solid plan.
I think we're pretty close tothat 95% clarity.
(14:55):
The only other thing I mightwant to know, just to finalize
everything, is if you have anyspecific requirements for how
the chatbot should respond.
For example, do you want it togive a summary of customer
sentiment or just list out thecall or offer a combination of
both?
Once we know that, we can puttogether a nice infrastructure
document and you'll be good togo.
SPEAKER_01 (15:13):
I'd like the chatbot to respond pretty directly with
specific feedback based on thequery from the product team,
which also means that I wouldideally want to strip some of
the call transcripts of any kindof small talk or personal
details and things like that aswell.
So I would like your input onhow we could potentially do that
(15:35):
in an automated fashion becauseit's unrealistic for us to go
and scrape hundreds and hundredsof call transcripts.
SPEAKER_00 (16:17):
Absolutely.
That all makes perfect sense.
And I love the focus on keepingit concise, but also allowing
deep links for more context.
So let's break that down intothe final pieces.
First, on the automatedstripping of small talk or
personal details.
One approach is to use a naturallanguage processing step in your
automation.
For example, you could have astep in Zapier that runs the
transcript through an AI textcleaner.
(16:39):
There are some pre-built toolsor even just a simple custom GPT
model that can identify andremove filler conversation.
That way you're only keeping thecore feedback.
We can definitely outline thatin the infrastructure doc.
Second, on the output side, it'stotally doable to have the
chatbot give a concise summaryand then provide a link directly
to the relevant call in Gong.
That way, product managers canclick through if they want to
(17:00):
hear the full context.
We'll just set up the chatbotresponses so they have a short
summary plus a listen to thefull call link.
With those final pieces inplace, I think we have
everything we need.
I'll go ahead and draft out aninfrastructure document that
covers the entire flow, pullingdata from Gong, cleaning it up,
storing it in Airtable, and thenusing Zapier to power the
chatbot interface.
If that all sounds good, we'llget that ready for you.
SPEAKER_01 (17:21):
That sounds good.
When you create this document,um ideally please um provide it
in PDF format.
And I would like for you to beas specific as possible when it
comes to the setup of each step.
So for example, in Airtable, Iwould want you to be very
specific in terms of whichcolumns we need to set up for
the data being pulled.
(17:41):
For instance, a link to, youknow, um the call transcript
itself and those kinds ofthings.
So be very specific in thosesteps and essentially treat it
as if you're teaching a beginnerhow to set up this kind of
automation with each tool.
SPEAKER_00 (17:57):
Absolutely.
I'll make it asbeginner-friendly and detailed
as possible.
I'll include all the specificsfor each step, like how to set
up your air table columns, whereto link the call transcripts,
and how to configure each partof the Zapier flow so that it's
all crystal clear and a nice PDFfor you.
So I'll put that together andwe'll have you all set up big
go.
SPEAKER_01 (18:14):
Great.
Thank you.
SPEAKER_00 (18:16):
You're very welcome.
I'm glad we could get everythingironed out.
I'll get that detailed PDF readyfor you with all the
step-by-step instructions.
And if you have any morequestions along the way, just
let me know.
Happy to help.
SPEAKER_01 (18:29):
Okay, so that is essentially the conversation I
had with it.
Again, I will put the outputs,that PDF output of what it came
back with in the show notesbelow so that if you want to go
look at it, you definitely can.
Uh again, would love yourfeedback and would love to hear
from you what kinds of thingsyou're chatting with ChatGPT
(18:50):
about on a regular basis.
But that'll do it for thisepisode.
I want to thank you for joiningme today and uh hope your week
ahead is an awesome one.
We'll see you next week.
Thank you for joining me forthis episode of the Digital CX
Podcast.
If you like what we're doing, uhconsider leaving us a review on
your podcast platform of choice.
If you're watching on YouTube,leave a comment down below.
(19:12):
It really helps us to grow andprovide value to a broader
audience and get moreinformation about the show and
some of the other things thatwe're doing at digitalcustomer
success.com.
I'm Alex Turkovich.
Thanks so much for listening,and we'll talk to you next week.
Popular Podcasts
Las Culturistas with Matt Rogers and Bowen Yang
Ding dong! Join your culture consultants, Matt Rogers and Bowen Yang, on an unforgettable journey into the beating heart of CULTURE. Alongside sizzling special guests, they GET INTO the hottest pop-culture moments of the day and the formative cultural experiences that turned them into Culturistas. Produced by the Big Money Players Network and iHeartRadio.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.