3011: AI, Data Privacy, and the Future of Marketing: Insights from Constellation's CTO

[00:00:03] [SPEAKER_02]: Welcome back to the Tech Talks Daily Podcast. Quick question, how is AI reshaping the way major

[00:00:12] [SPEAKER_02]: corporations engage with their audiences and protect their data and so much more in between?

[00:00:19] [SPEAKER_02]: Well, in today's episode my guest is a tech and data expert. He currently serves as a

[00:00:27] [SPEAKER_02]: an all-round great guy and a lot of fun to be around. And today he's going to be also bringing

[00:00:33] [SPEAKER_02]: with him an understanding of the intersection between AI, data security, and innovative

[00:00:39] [SPEAKER_02]: marketing strategies. But at the heart of our conversation today is Constellations proprietary

[00:00:46] [SPEAKER_02]: technology which consists of a groundbreaking AI tool that leverages contextual social listening

[00:00:53] [SPEAKER_02]: to elevate the marketing efforts of Fortune 100 companies. But what does that mean? It means

[00:00:59] [SPEAKER_02]: driving revenue and enhancing brand engagement. So my guest today will walk me through how

[00:01:06] [SPEAKER_02]: these unique capabilities are setting a new standard in the industry. And we'll also dive into

[00:01:13] [SPEAKER_02]: the broader AI landscape, explore the strides that tech giants like the usual suspects

[00:01:18] [SPEAKER_02]: Google, Microsoft and Apple are making. And finally in an age where AI is becoming ubiquitous

[00:01:25] [SPEAKER_02]: how can companies safeguard their data? Well my guest today is also going to provide expert

[00:01:30] [SPEAKER_02]: insights into the challenges and best practices for maintaining data security

[00:01:36] [SPEAKER_02]: in a world increasingly influenced by AI. I want to take a time out to express my

[00:01:42] [SPEAKER_02]: gratitude to everyone who supports my mission of delivering daily content to you in 165 countries.

[00:01:48] [SPEAKER_02]: I couldn't do it without you and I couldn't do it without my sponsors. And today I want to give

[00:01:53] [SPEAKER_02]: a quick shout out to Kiteworks who recently told me that defence contractors are facing immense

[00:01:59] [SPEAKER_02]: pressure to comply with things like CMMC 2.0 security standards and finding a secure easy

[00:02:05] [SPEAKER_02]: to use file sharing platform that meets those guidelines can be a big challenge.

[00:02:11] [SPEAKER_02]: So quick shout out to any defence contractors listening out there. CMMC 2.0 compliance doesn't

[00:02:18] [SPEAKER_02]: have to be a headache. Consider Kiteworks your fast track to authorization and as a

[00:02:23] [SPEAKER_02]: FedRAMP moderate authorised solution they cover nearly 90% of CMMC 2.0 level 3 requirements.

[00:02:30] [SPEAKER_02]: For you that means less time, less effort, less cost and while others struggle with DIY

[00:02:35] [SPEAKER_02]: approval processes and clunky apps you'll breeze through with their Zero Trust frameworks.

[00:02:41] [SPEAKER_02]: So don't let compliance slow you down is what I'm trying to say. Simply visit

[00:02:45] [SPEAKER_02]: kiteworks.com to get started. That's kiteworks.com to learn more about Kiteworks' secure

[00:02:51] [SPEAKER_02]: content platform for CMMC compliance. But with my thank yous out the way it's now time to

[00:02:57] [SPEAKER_02]: jump right into today's interview with a fantastic guest.

[00:03:02] [SPEAKER_02]: So are you ready to explore the future of AI and its impact on businesses? Well let's dive in.

[00:03:09] [SPEAKER_02]: So buckle up and hold on tight as I beam your ears all the way to New York where my guest

[00:03:15] [SPEAKER_02]: is waiting to join me today. So a massive warm welcome to the show. Can you tell everyone

[00:03:22] [SPEAKER_02]: listening a little bit who you are and what you do? Yeah I appreciate it Neil it's really exciting to

[00:03:28] [SPEAKER_01]: be here. My name is Nomanafiz I'm the CTO at Constellation. I have been with Constellation for

[00:03:36] [SPEAKER_01]: seven years and before that I've just been in the product tech solutions and development area

[00:03:43] [SPEAKER_01]: for you know 15 plus years before that and really I started my career working at RGA which

[00:03:50] [SPEAKER_01]: is a product shop and we work closely with Nike and Samsung and Disney and what we were doing for them

[00:03:56] [SPEAKER_01]: and this is like you know in 2007, 8, 9, 10 and post that time so it's really kind of building

[00:04:04] [SPEAKER_01]: some of those first iterations of the web the more interactive web the more scalable web and

[00:04:09] [SPEAKER_01]: it's like so we you know we worked with Nike.com Nike store a lot of those properties

[00:04:14] [SPEAKER_01]: at Nike was kind of you know starting to create a digital presence and a lot of product work

[00:04:19] [SPEAKER_01]: for them. The last project I worked at with RGA was Marvel.com so we rebuilt Marvel.com from the

[00:04:26] [SPEAKER_01]: ground up you know content heavy like you know it's a huge destination for a lot of fans and

[00:04:31] [SPEAKER_01]: such amazing content in there but it's a huge engineering exercise beneath that and you

[00:04:37] [SPEAKER_01]: know just scale of the type of environment that they have so that was really exciting and

[00:04:42] [SPEAKER_01]: I was essentially the global infrastructure lead for RGA and then I joined Constellation

[00:04:48] [SPEAKER_01]: to just you know own all of our tech and data and IT and that's what I've been doing for the last

[00:04:55] [SPEAKER_01]: seven years and at Constellation I oversee you know our engineering data and product team

[00:04:59] [SPEAKER_01]: is how I like to see it and we're really just building a couple different systems. We have a

[00:05:06] [SPEAKER_01]: SaaS platform that's primarily focused on the compliance and regulatory verticals so we work

[00:05:12] [SPEAKER_01]: primarily in auto insurance and finance and pharma and what our solution does is not only does it do

[00:05:20] [SPEAKER_01]: a lot of analysis on the data side but it uses that analysis to actually drive content creation

[00:05:26] [SPEAKER_01]: and all of those things at huge scale. So it's about unblocking data teams as well as

[00:05:33] [SPEAKER_01]: strategy teams as well as content creation and marketing teams so that they can really

[00:05:38] [SPEAKER_01]: create content that's compliant ready to go out and doesn't take you know days or months or even

[00:05:44] [SPEAKER_01]: weeks sometimes within some of these industries to actually get approved content.

[00:05:50] [SPEAKER_02]: Well I'm so happy to hear you join me on the podcast today because one of the things I

[00:05:54] [SPEAKER_02]: always try and do on this show every day is demystify technology and especially in a world where

[00:06:00] [SPEAKER_02]: everyone's gone crazy for AI and there's a lot of business leaders in Fortune 500 companies thinking

[00:06:06] [SPEAKER_02]: well what does this mean for me and our business and what does this mean for my customers and how

[00:06:10] [SPEAKER_02]: we're going to use it and all that stuff so many questions and then I come across Constellation

[00:06:15] [SPEAKER_02]: and your proprietary technology leverages AI to enhance contextual social listening

[00:06:21] [SPEAKER_02]: for Fortune 100 companies. I thought well this is just perfect for the show so

[00:06:25] [SPEAKER_02]: tell me a bit more about this and what you're doing here.

[00:06:29] [SPEAKER_01]: Yeah no it's really interesting I mean you know what we found working with some of these large

[00:06:33] [SPEAKER_01]: businesses is that there's a few ways that AI can help them and it's really about you know

[00:06:41] [SPEAKER_01]: they have so much data and so much you know sort of proprietary analysis that you have to unpack

[00:06:49] [SPEAKER_01]: and unlock and be able to articulate into something actionable so you know our system does partly it

[00:06:57] [SPEAKER_01]: looks at a lot of public data sets so if you think of the automotive industry and those types of

[00:07:02] [SPEAKER_01]: verticals right there's a lot of data whether it's vehicle inventory whether it's customer reviews

[00:07:09] [SPEAKER_01]: whether it's you know specific incentives that OEMs put out that is usually a big bottleneck

[00:07:17] [SPEAKER_01]: it's just being able to pull in that type of data and organize it and structure it in a way that you

[00:07:22] [SPEAKER_01]: can you can leverage it so we do that type of data ingestion and data kind of organization

[00:07:31] [SPEAKER_01]: and that's just kind of the more public data set so that's where AI is used you know in very

[00:07:38] [SPEAKER_01]: specific layers of the stack to whether it's summarize or extract or figure out you know what

[00:07:48] [SPEAKER_01]: an insight is for a very specific kind of piece of data and I would say kind of the what I'd like

[00:07:56] [SPEAKER_01]: to think about in that to answer your question is AI has a small piece of this larger sort of

[00:08:02] [SPEAKER_01]: stack of decision-making and that's really where it can be useful you know a lot of people

[00:08:05] [SPEAKER_01]: think of oh I'm just gonna like you know throw AI at these big problems and expect to get like an

[00:08:11] [SPEAKER_01]: amazing solution and get kind of you know something that's just going to take me to the finish line

[00:08:14] [SPEAKER_01]: and that's really not how it pans out especially for Fortune 100 and those types of companies

[00:08:19] [SPEAKER_01]: it's you really kind of have a whole you know stack of data sets and decisions that need to be

[00:08:26] [SPEAKER_01]: made and you can apply AI individually to a few of them and that's how you start to kind of

[00:08:31] [SPEAKER_01]: slowly shorten the gap with some of those very specialized kind of workflows and that's really

[00:08:37] [SPEAKER_01]: that's really where we are focused we're not using AI to just give you insights but we're

[00:08:43] [SPEAKER_01]: getting using AI to get you from A to Z faster and it's a really iterative process to continue

[00:08:49] [SPEAKER_02]: to unpack how we leverage it. I think there's so many different businesses wanting to be a part

[00:08:56] [SPEAKER_02]: of that AI narrative at the moment but also if we look I think every single tech project in every

[00:09:02] [SPEAKER_02]: sector in every organization people get a little bit wise to some of those expensive IT projects that

[00:09:08] [SPEAKER_02]: don't always deliver and don't always get delivered on time so now people are starting to look at

[00:09:14] [SPEAKER_02]: what's the ROI, what's the business value, what's the measurable difference and what's the tangible

[00:09:19] [SPEAKER_02]: results we can expect so can you tell me a little bit more about the impact that your

[00:09:24] [SPEAKER_02]: technology has had on some of the marketing strategies revenue growth of some of the biggest

[00:09:29] [SPEAKER_01]: brands in the world? Yeah, you know, a hundred percent. I mean, you know, it's really about what

[00:09:35] [SPEAKER_01]: art tech is used for is first of all once you even have an understanding of how you want to

[00:09:44] [SPEAKER_01]: market your specific campaigns or how you want to reach out to your audience

[00:09:52] [SPEAKER_01]: a lot of times these companies have trouble getting through the compliance process. You know,

[00:09:57] [SPEAKER_01]: they have a lot of lawyers, everyone's like really hesitant to you know say okay you can create a

[00:10:05] [SPEAKER_01]: hundred thousand personalized pieces of content because there's a review process in there, there's

[00:10:10] [SPEAKER_01]: a number of review processes in there and they just don't have the capability to kind of work

[00:10:16] [SPEAKER_01]: at that type of scale and that's where first of all what our software does is it actually

[00:10:21] [SPEAKER_01]: generates all the skews of content that you need from data. So it uses data mixed with templates

[00:10:29] [SPEAKER_01]: and other sort of rules that are built into the system to just generate all your content

[00:10:34] [SPEAKER_01]: upfront so that's a first of all a huge time saving that's normally something where

[00:10:39] [SPEAKER_01]: that's manual work that either an agency or someone at the team is doing and it's really

[00:10:46] [SPEAKER_01]: error prone and it's really slow in terms of actually kind of making that content. So that's

[00:10:52] [SPEAKER_01]: the first part that we use our system to do and then from there we actually do a lot of

[00:11:00] [SPEAKER_01]: approval and review within our platform so we structure content in a way that even if you

[00:11:08] [SPEAKER_01]: have to you review 100,000 pieces of content you understand the building blocks and you

[00:11:13] [SPEAKER_01]: understand what's changing so if there is some sort of review and compliance process

[00:11:19] [SPEAKER_01]: they don't have to review every item individually they just need to look at that bigger picture and

[00:11:23] [SPEAKER_01]: see what are some of the major pieces that are changing as I go through this content. So

[00:11:27] [SPEAKER_01]: that's like a huge KPI improvement you know and it allows companies to start to think

[00:11:32] [SPEAKER_01]: strategically about okay I can reach my audience at a more granular level I don't need to do

[00:11:38] [SPEAKER_01]: this broad type of marketing or broad type of sales initiatives I can actually say okay for every

[00:11:44] [SPEAKER_01]: you know person who's in a specific set of zip codes I can give them something that

[00:11:49] [SPEAKER_01]: is a little bit more personal and kind of is reflective of their interests and you can kind

[00:11:54] [SPEAKER_01]: of take that and continue to kind of move it forward. And as quickly as this pace of

[00:12:00] [SPEAKER_02]: technology seems to be moving at the moment there's also a realization it might not move

[00:12:04] [SPEAKER_02]: this slower game there's always talk of AGI by 2029 so as AI continues to advance

[00:12:12] [SPEAKER_02]: are there any key developments that you see from all the usual suspects such as Google,

[00:12:17] [SPEAKER_02]: Microsoft, Apple and Co in the AI space anything that's got your attention or excites you at the

[00:12:23] [SPEAKER_01]: moment? Yeah I mean it's really interesting how you know it's become like front and center for

[00:12:28] [SPEAKER_01]: every one of these companies I mean they're all sort of pivoted you could say in a way I

[00:12:32] [SPEAKER_01]: meta changed their name and then we have to kind of totally switch focus Google felt like they were

[00:12:39] [SPEAKER_01]: faltering but even they're all constantly kind of highlighting how unpredictable AI can be. I mean

[00:12:48] [SPEAKER_01]: Google Gemini they just had their big kind of reveal of some of their new devices and

[00:12:54] [SPEAKER_01]: Gemini failed like a couple of times on stage right there and then which is like the biggest

[00:13:00] [SPEAKER_01]: you know I mean that's a massive sort of highlight to like how challenging these systems can be to

[00:13:07] [SPEAKER_01]: actually get right and that's more like a scale problem than it is an intelligence problem even

[00:13:13] [SPEAKER_01]: but really what excites me to be honest is personally kind of speaking from a developer

[00:13:18] [SPEAKER_01]: standpoint I think what these tools do more than anything is allow you as a developer to

[00:13:27] [SPEAKER_01]: focus on the things that are more you know specific to your product in your stack and allow

[00:13:35] [SPEAKER_01]: you to kind of move a little bit faster whether it's writing unit tests or writing you know sort of

[00:13:39] [SPEAKER_01]: reviewing a function and seeing if there's some edge cases and obviously you have to continue to

[00:13:46] [SPEAKER_01]: you know look at the answers and sort of dissect them you can't trust everything that these

[00:13:51] [SPEAKER_01]: systems tell you but they are an amazing way to get a first glance and kind of get a first take

[00:13:58] [SPEAKER_01]: on something that you're working on and kind of become like a handy assistant and I think they

[00:14:03] [SPEAKER_01]: work really well for developer type of tasks initially I think we're going to see that you

[00:14:08] [SPEAKER_01]: know scale into many different workflows but but I really feel that they are just another

[00:14:14] [SPEAKER_01]: tool your toolbox that allows you to move a little bit faster and a little bit more confidently

[00:14:18] [SPEAKER_01]: but where they you have to be careful is when someone who is not an expert in the field is tapping

[00:14:25] [SPEAKER_01]: into those systems because that's where you tend to trust them more than they you know are normally

[00:14:33] [SPEAKER_01]: you know valuable for and you they tend to give you answers with the level of confidence that can

[00:14:38] [SPEAKER_01]: you know seem a little bit reliable but you kind of need to take that with a grain of salt

[00:14:43] [SPEAKER_01]: sometimes and really kind of review the answers but I do think I'm really excited about how like

[00:14:48] [SPEAKER_01]: the day-to-day for some of these roles is just more focused on the things that you care about and

[00:14:53] [SPEAKER_01]: less focused on some of the manual grind that ends up happening in any role.

[00:14:58] [SPEAKER_02]: And outside of those magnificent seven I've lost count of how many tech conferences I've been

[00:15:03] [SPEAKER_02]: to every single keynote begins with this is our AI solution but we've been working on this

[00:15:09] [SPEAKER_02]: for 10 years before anybody else yeah and that but now we're embracing the copilot thing and

[00:15:15] [SPEAKER_02]: what we're gonna do it ethically and responsibly blah blah blah blah blah.

[00:15:19] [SPEAKER_02]: How do you envision all these other companies responding to or following the lead of these

[00:15:24] [SPEAKER_01]: major players in AI? What do you see here? Yeah I mean it's interesting I mean everyone's sort

[00:15:32] [SPEAKER_01]: of got their story now it feels like I mean Walmart's like oh we saved so much time using

[00:15:37] [SPEAKER_01]: General AI and like everyone's trying to spin you know whatever they're doing to incorporate

[00:15:43] [SPEAKER_01]: AI in some way you know I really think there's gonna there's sort of these ups and down

[00:15:47] [SPEAKER_01]: hype curves that we always see and I feel like you know we kind of get really excited about

[00:15:54] [SPEAKER_01]: you know GPT-4, 4L comes out and everyone sort of like surf in that wave and then you start

[00:15:59] [SPEAKER_01]: to unpack okay it's actually not as smart and some things as the original GPT-4 you know

[00:16:06] [SPEAKER_01]: and so there's always that up and down and I feel like every team needs to just be a little bit

[00:16:12] [SPEAKER_01]: you know self-aware of how they're sort of relying on these systems and trusting them

[00:16:17] [SPEAKER_01]: and you know I see the magic in what they do but you really need to kind of take that with

[00:16:23] [SPEAKER_01]: a grain of salt and start to really it's kind of like the original kind of how I was talking

[00:16:28] [SPEAKER_01]: about it like you might have a whole stack of problems you're trying to solve and there's

[00:16:31] [SPEAKER_01]: certain pieces of that that I think yeah I can do really well and you need to make sure that you're

[00:16:37] [SPEAKER_01]: taking that sort of approach with every way you're tapping into AI and sort of

[00:16:43] [SPEAKER_01]: seeing okay these are the areas where I can kind of have it put to use and put to good use

[00:16:48] [SPEAKER_01]: and these are the areas where I always need to make sure that you know someone's checking

[00:16:51] [SPEAKER_01]: the work and someone's kind of bringing the work and so for all of these types of companies

[00:16:56] [SPEAKER_01]: at scale you know I think there's this rush to sort of talk about these things I almost feel like

[00:17:03] [SPEAKER_01]: they should slow down a little bit put it to work see how their results are actually panning out

[00:17:09] [SPEAKER_01]: and then talk about those case studies as opposed to like you know putting AI in front of every

[00:17:13] [SPEAKER_01]: product term on their website and just sort of saying like it's AI enabled like it's easy to

[00:17:18] [SPEAKER_01]: just make an API request and like say like okay it's AI enabled now but there's no like

[00:17:23] [SPEAKER_01]: actual benefit apart it becomes like a toy a little bit it's like okay that's fun but is it

[00:17:29] [SPEAKER_01]: actually making my work better and making me work faster that's always a little bit of the you know

[00:17:34] [SPEAKER_02]: those details 100% with you and with AI playing an increasingly significant role in data processing

[00:17:41] [SPEAKER_02]: as well are any strategies that you think companies should be adopting to maybe ensure

[00:17:48] [SPEAKER_02]: that their data remains secure because it's very easy to throw everything into that black box

[00:17:52] [SPEAKER_02]: and everything's excited but securing when you talk about that is a point right no you really do

[00:17:58] [SPEAKER_01]: I mean it's so interesting because like we've been you know playing around with using AI to do data

[00:18:05] [SPEAKER_01]: retrieval and data analysis a lot like I was saying and one thing that you start to see is

[00:18:10] [SPEAKER_01]: it's also really easy for people to ask weird questions and then depending on what AI has

[00:18:17] [SPEAKER_01]: access to it'll it's more than welcome to like you know show what your root directory contains or

[00:18:23] [SPEAKER_01]: show what your environment variables are and things like that so you really need it's almost like

[00:18:30] [SPEAKER_01]: more work needs to be spent on the actual you know validation of data validation of answers

[00:18:37] [SPEAKER_01]: validation of the prompt itself and whether it's something that can be trusted whether

[00:18:41] [SPEAKER_01]: it's something that's you know context relevant so like it's almost like you really need to be

[00:18:47] [SPEAKER_01]: careful of how the logical system in these EIS is working and you know whether the user's asking

[00:18:54] [SPEAKER_01]: for something that makes sense or they're trying to hack the system in some ways like

[00:18:59] [SPEAKER_01]: it's very easy to say like okay I'm going to give it access to this SQL table and then

[00:19:03] [SPEAKER_01]: it's going to do something cool and then just like give the user anything they want but

[00:19:07] [SPEAKER_01]: permissions and security even in that relatively straightforward workflow

[00:19:11] [SPEAKER_01]: are always like the most important thing and you can see like we we did test initially where

[00:19:17] [SPEAKER_01]: yeah if you would ask one of these systems something in a very specific way it would actually kind of

[00:19:22] [SPEAKER_01]: reveal you know like specific environment variables like the EPI keys and other things

[00:19:26] [SPEAKER_01]: that you know are just accessible to it because it's running on a specific environment

[00:19:32] [SPEAKER_01]: so again you know it's something that you have to be forefront and you have to test every little

[00:19:38] [SPEAKER_01]: piece of this and it's really but also then building automation that tests ways the tests

[00:19:45] [SPEAKER_01]: your your sort of model in hundreds and hundreds of different ways and it's kind of going back

[00:19:50] [SPEAKER_01]: to that you know Netflix had that what was it called the chaos monkeys where they are

[00:19:55] [SPEAKER_01]: basically this is I'm probably a little while ago now but they had like a very randomized set of

[00:20:01] [SPEAKER_01]: system that would randomly take down services and sort of like cause chaos

[00:20:06] [SPEAKER_01]: and we also need to use a little bit of that philosophy with how we test the AI because

[00:20:10] [SPEAKER_01]: you might think okay this is what it's supposed to do I'm going to test it in this way

[00:20:15] [SPEAKER_01]: but it's all the stuff that you're not thinking about where it's actually going to

[00:20:19] [SPEAKER_02]: potentially you know fail but yeah so if we were to compare your AI driven tools in the

[00:20:26] [SPEAKER_02]: market especially in terms of things like data security contextual accuracy all the things

[00:20:31] [SPEAKER_02]: we're talking about with other solutions out there what is it that you think makes your

[00:20:36] [SPEAKER_02]: stand out what makes it different from all these other solutions yeah it's a great question one thing

[00:20:42] [SPEAKER_01]: we really focus on and pride ourselves on is the accuracy of the results that our data

[00:20:48] [SPEAKER_01]: systems are responding to that's the first part and the second one is the the industry's

[00:20:55] [SPEAKER_01]: understanding so if you take automotive for example there's many different ways that

[00:21:00] [SPEAKER_01]: you know you can ask a question but it's this automotive lingo the specific lingo that you

[00:21:06] [SPEAKER_01]: know whether it's automotive dealerships or oems or even agencies use and so our systems understand

[00:21:13] [SPEAKER_01]: that and then they also understand the data sets that are very specific to automotive in that example

[00:21:20] [SPEAKER_01]: and they are tailored to give accurate results or give no results at all so we'd rather

[00:21:29] [SPEAKER_01]: you know skew to accuracy and it's like you know kind of tailor everything and test everything to

[00:21:34] [SPEAKER_01]: meet that and if it doesn't know the answer because as you've probably seen sometimes you can ask

[00:21:40] [SPEAKER_01]: these AS systems question and they just love giving you an answer so they will just make something

[00:21:44] [SPEAKER_01]: up you know and it's like yep this is exactly what it is but we would much rather that the system

[00:21:51] [SPEAKER_01]: just says I don't know the answer or just gives you like you know sort of something

[00:21:56] [SPEAKER_01]: that is guiding guiding but not an actual answer so I would say like those are definitely the

[00:22:01] [SPEAKER_01]: key areas that our AI systems are focused on 100 with you I've got a great example of that

[00:22:07] [SPEAKER_02]: I've just pulled up there because I was doing a bit of research on an article about a company

[00:22:12] [SPEAKER_02]: and how many users they had and the response that this was clawed AI that told me this as of late 2023

[00:22:19] [SPEAKER_02]: the company had 50 million users worldwide so my response to that was okay that's brilliant

[00:22:26] [SPEAKER_02]: and but can you back up that figure with a stat and with the source from that stat

[00:22:32] [SPEAKER_02]: and this is the exact response that it gave me Neil I apologize for the confusion in my previous

[00:22:38] [SPEAKER_02]: response upon reflection I realize I don't have a source for the specific amazing

[00:22:44] [SPEAKER_00]: love it yeah exactly it's like but I love how it phrased that even upon reflection it realized

[00:22:52] [SPEAKER_02]: oh man well I mean for yourself I mean drawing from your experience at constellation and your

[00:22:58] [SPEAKER_02]: work with major brands what are the biggest challenges companies are facing when integrating AI

[00:23:03] [SPEAKER_02]: into their operations you must be the heir of so many different big companies what are

[00:23:09] [SPEAKER_00]: their biggest challenges what are they talking about yeah I mean it's uh yeah it's again like

[00:23:15] [SPEAKER_01]: everyone's excited about AI but they're also just unsure how to apply it and you know what

[00:23:22] [SPEAKER_01]: we kind of work with them on is taking the things that um you know really well so I think one

[00:23:31] [SPEAKER_01]: reason why constellations been very successful is we've done all of the tasks from A to Z in a

[00:23:39] [SPEAKER_01]: manual way or in some sort of way before we've automated them so we understand the intricacy

[00:23:46] [SPEAKER_01]: of a workflow and then we understand exactly how to automate it and how to apply AI to it and I

[00:23:52] [SPEAKER_01]: think that sort of philosophy is how a lot of these companies should also sort of work it's

[00:23:57] [SPEAKER_01]: yep you've been taking this long-winded path to get a process done or to get a workflow done and

[00:24:02] [SPEAKER_01]: you know those are the processes which you should start to look at okay where can I apply AI or

[00:24:08] [SPEAKER_01]: where can I apply automation where can I apply some of these new modernizing solutions because

[00:24:15] [SPEAKER_01]: you know exactly how to test those systems you know exactly how to validate them you know

[00:24:20] [SPEAKER_01]: exactly how to you know sort of ensure that you're getting to the same point and the KPIs

[00:24:26] [SPEAKER_01]: become you know speed and quality and all the other things and it's it's much easier to kind of work

[00:24:31] [SPEAKER_01]: through that than it is to try to build a new product and apply AI to it and try to solve all

[00:24:37] [SPEAKER_01]: these things at the same time I feel like that you're opening up a lot of questions and you're

[00:24:42] [SPEAKER_01]: kind of you know setting yourself up in uh in a way that yeah high high chance of failure so

[00:24:48] [SPEAKER_01]: it's really kind of how we've sort of iterated and sort of evolved from like okay we have a

[00:24:55] [SPEAKER_01]: tech platform that does all these things and now we want to reduce that time significantly

[00:24:59] [SPEAKER_01]: but do it just as well if not better so that would be my kind of high level uh direction.

[00:25:06] [SPEAKER_02]: Love that a big question for you here because I know everything that we're talking about

[00:25:10] [SPEAKER_02]: is such a fine delicate balance so how do you balance innovation in one hand with security

[00:25:16] [SPEAKER_02]: when developing tech solutions that need to scale across such diverse industries that

[00:25:22] [SPEAKER_02]: we make it sound so simple when we're talking about it on podcasts or reading about it online

[00:25:27] [SPEAKER_01]: of course it is so complex how do you balance it? Yep no it's you know testing it comes down to like

[00:25:35] [SPEAKER_01]: your uh like testing should be the responsibility of everyone on the team I think that's really

[00:25:42] [SPEAKER_01]: what it you know like people like developers don't want to write unit tests right automation tests

[00:25:48] [SPEAKER_01]: and like project managers don't want to test they just like leave it all in the QA team with a

[00:25:52] [SPEAKER_01]: security team or whatever it is it's really up to every member of the team to be testing

[00:25:58] [SPEAKER_01]: and to be checking all parts of the platform or product whatever they're aware of and and

[00:26:05] [SPEAKER_01]: they have understanding of and it really is about becoming smarter in terms of how we write our

[00:26:11] [SPEAKER_01]: tests and how we write our automation because that's I feel like a part of the system that is always

[00:26:19] [SPEAKER_01]: you know that's where you have the least amount of time because that's when things actually mature

[00:26:23] [SPEAKER_01]: enough so you can test them and then you're meeting you're trying to meet deadlines you're

[00:26:28] [SPEAKER_01]: trying to make whatever launch dates etc and that's the first thing that sort of gets cut you

[00:26:32] [SPEAKER_01]: know it's like okay the testing the automation and the actual you know security arduing and

[00:26:39] [SPEAKER_01]: I mean you see with CrowdStrike you see with all the systems I mean at the end of the day like

[00:26:44] [SPEAKER_01]: it's the lack of testing in the right environments and in the right specific scenarios that lead to

[00:26:50] [SPEAKER_01]: those sorts of massive failures and we as larger organizations we need to just spend much more

[00:26:57] [SPEAKER_01]: of our resources on testing and quality and it comes all the way down to the code level

[00:27:03] [SPEAKER_01]: and to the infrastructure and to every part of the organization that's how you know

[00:27:08] [SPEAKER_01]: it really needs to be kind of the bread and butter of any of these students

[00:27:12] [SPEAKER_02]: and I think that's a beautiful moment to end on I can't thank you enough for coming on and

[00:27:16] [SPEAKER_02]: sharing your insights today but before I let you go I'm gonna ask you to leave one final gift for

[00:27:21] [SPEAKER_02]: everyone listening I always ask my guests to leave either a book that means something to them

[00:27:26] [SPEAKER_02]: we can recommend on our Amazon wishlist or a song for our tech talks daily spot if I play

[00:27:33] [SPEAKER_02]: like skillet pleasures are allowed but what would you like to leave them?

[00:27:37] [SPEAKER_01]: Yeah it's funny there was a song I don't even know where I first heard this but I heard this

[00:27:42] [SPEAKER_01]: long time ago and then it came up again and I really really love it it's by uh Joan Baez

[00:27:50] [SPEAKER_01]: it's called Diamonds and Rust it's a it's a really interesting song because you it's

[00:27:56] [SPEAKER_01]: from my understanding it's actually a story of her relationship with Bob Dylan

[00:28:02] [SPEAKER_01]: and it sort of is a really beautiful tale of sort of you know a time in her life it's just an

[00:28:11] [SPEAKER_01]: amazing song and it sort of just came back to me I don't even know how I heard it again but it really

[00:28:16] [SPEAKER_01]: is something that uh yeah just kind of blew my mind again and the way the tune is in the way

[00:28:23] [SPEAKER_01]: songs so anyway I'm gonna put on a rabbit hole a little bit but yeah I think it's a great

[00:28:26] [SPEAKER_02]: addition to your playlist. Wow if it's good enough for you it's good enough for me and

[00:28:30] [SPEAKER_02]: Bobby McGee well that's another song for me right now for another day but I knew you were a

[00:28:35] [SPEAKER_02]: classy guy with that you'd have a great tune lined up for me an absolute banger so I will get that

[00:28:40] [SPEAKER_02]: added to a Spotify playlist and for anyone listening just want to find out more information about

[00:28:45] [SPEAKER_02]: you you'll connect with your team or or just dig deeper on any of the tech and the stuff

[00:28:50] [SPEAKER_02]: that you're doing here where would you like to point everyone? Yeah I mean you know our

[00:28:56] [SPEAKER_01]: website hellocostlation.com is a great place to find all of our channels I'm not I don't have a huge

[00:29:02] [SPEAKER_01]: presence on any channel personally but you know I'm always sort of around and you can sort of find

[00:29:11] [SPEAKER_01]: me through the website and other methods. Well love Chai with you today learning more about

[00:29:17] [SPEAKER_02]: Constellations proprietary technology how it uses AI to perform contextual social listening

[00:29:23] [SPEAKER_02]: improving marketing capabilities and revenue for so many Fortune 100 companies that is huge on its

[00:29:31] [SPEAKER_02]: own but I also love that delicate balance we're talking about about keeping data safe with AI in

[00:29:37] [SPEAKER_02]: the picture now I think we could talk for ages on this to get you back on next year maybe see how

[00:29:43] [SPEAKER_02]: things are evolving but thank you for sharing your story today. Amazing I appreciate that Neil it's

[00:29:48] [SPEAKER_02]: great to be on. So as we close today's episode I think one thing is clear AI is not just transforming

[00:29:56] [SPEAKER_02]: how companies market to their audiences it's redefining an entire business landscape and my

[00:30:02] [SPEAKER_02]: guest today offered a little glimpse into the future of marketing a future where AI driven

[00:30:08] [SPEAKER_02]: contextual listening becomes a powerful tool for enhancing customer engagement and driving revenue

[00:30:16] [SPEAKER_02]: real tangible differences real metrics behind them as well but with great power comes great

[00:30:23] [SPEAKER_02]: but responsibility I've watched enough Spider-Man in my life to know all about that and as AI

[00:30:29] [SPEAKER_02]: continues to evolve so too do the challenges of keeping that data safe and I cannot thank my

[00:30:35] [SPEAKER_02]: guest enough for highlighting the need for companies to stay ahead of the curve by

[00:30:39] [SPEAKER_02]: adopting robust security measures that account for the complexities that AI introduces

[00:30:46] [SPEAKER_02]: so what steps will your organization take to harness the power of AI while safeguarding its data

[00:30:54] [SPEAKER_02]: reflecting on our conversation today how could some of the things we talk about be applied

[00:30:59] [SPEAKER_02]: to your business as always email me tech blog writer outlook.com twitter linkedin instagram

[00:31:06] [SPEAKER_02]: just at neil cqs let me know your thoughts other than that thank you for joining me today on on this

[00:31:13] [SPEAKER_02]: journey into the future of technology I'll be back again usual time same time same place but

[00:31:20] [SPEAKER_02]: until next time stay innovative stay secure and hopefully I will speak with you all again tomorrow