Latest / Tech Journeys Podcast / Deepankar Mathur | Building Secure AI Products While Running a Side Consultancy | The Consulting Launch Method Podcast
Transcript
- 0:00Welcome to the Tech Journey podcast.
- 0:02I'm your host, Vincent.
- 0:03And on this podcast, I interview people in tech.
- 0:06And always, this show is sponsored
- 0:08by the Consulting Launch Method,
- 0:10where I help experienced tech professionals
- 0:12build their own consulting business
- 0:14without spending years figuring out the business side
- 0:16alongside the day job.
- 0:18And we've helped clients go from inception
- 0:20to first flight in 90 days and six figures within a year.
- 0:23To learn more, connect with me on LinkedIn
- 0:25and send me a message.
- 0:28On today's podcast, we have Deep Mathur.
- 0:31And Deep works at Map Digital.
- 0:33And there, he's a product manager
- 0:35and he's focused around my tech,
- 0:38mobile engagements, software development kits,
- 0:41data integration, and customer-facing product development.
- 0:44So, Deep, welcome on the show.
- 0:47I always ask these questions first.
- 0:50Yeah, and I wanted to ask you this straight away.
- 0:53Looking back at your career,
- 0:56what's the decision or the project
- 0:58that you are most proud of and why?
- 1:02What really made it stand out for you?
- 1:05Well, thanks for having me on the podcast today, Vincent.
- 1:07I really appreciate it.
- 1:09Answering the question,
- 1:11there have been quite a few instances.
- 1:13My career journey hasn't been conventional or straight.
- 1:18I, in this journey, also had my own startup
- 1:22alongside my work.
- 1:23And I called it G10 Assist back in India.
- 1:26That taught me a lot of things.
- 1:28It taught me demand generation,
- 1:30taught me customer engagement.
- 1:32It also taught me how,
- 1:35what is the difference between having an idea
- 1:38to taking that idea into reality
- 1:40and building something versus building something sustainable.
- 1:44Because I had to eventually close it
- 1:46due to operational issues.
- 1:48I was able to generate demand,
- 1:50but the supply side was a problem,
- 1:52which was, I don't consider that as a failure,
- 1:54but a great learning lesson for myself.
- 1:57Because, yeah, as I said,
- 2:00it taught me sustainability
- 2:02versus just building something ground up.
- 2:05So that was a great learning lesson for me.
- 2:09And you say it's been conventional.
- 2:12And so that you go back to your sort of startup days,
- 2:16what is so unconventional about it?
- 2:18Because now, obviously, you're in a product management role,
- 2:23how did you make that transition
- 2:24or what steps got you there?
- 2:27Yeah, absolutely.
- 2:28That's a great question
- 2:29because I come from a background of technology
- 2:32and doing something hands-on.
- 2:35So I was a developer myself back in India.
- 2:38And then I moved to the UK,
- 2:40started working for Photobox
- 2:43as a marketing technology consultant.
- 2:45Put my hands on with the whole field
- 2:48of marketing technology.
- 2:49I understood the customer very well,
- 2:51or Photobox's customer.
- 2:53And then I was absorbed by MAP as a solutions architect.
- 2:58Now with MAP, in the solutions architect role,
- 3:01I had to dive a bit deeper into technology,
- 3:05understanding the customer better,
- 3:07not just one type of customer,
- 3:09not just one domain, but multiple domains.
- 3:12And as a problem solver,
- 3:13as basically a sitting CTO,
- 3:15so to say, from the MAP side,
- 3:19for a customer that I was working along,
- 3:22I had to understand their problems,
- 3:24their use cases, and how the value can be added.
- 3:27Now here's the most important learning of my life.
- 3:30Sometimes when we think technology is everything,
- 3:34I might be wrong.
- 3:35Because it is the customer's problem
- 3:39and that problem-solving aptitude,
- 3:41be it with technology or without technology,
- 3:44that at the end of the whole journey
- 3:47contributes to the most part.
- 3:49And this is where I started contributing
- 3:51to the product more,
- 3:52trying to solve, not trying to adjust what I have
- 3:55and solve the problem with what I have,
- 3:57but also contribute to making the product better itself.
- 4:01And then slowly I transitioned
- 4:02into the product management role,
- 4:04so my solutions role at MAP Digital itself.
- 4:07So now I am working as a product manager,
- 4:10solving the problems on a daily basis as I would,
- 4:14but just from the other perspective,
- 4:15from the customer's perspective in the product,
- 4:18rather than on the customer side,
- 4:21trying to solve the problem with whatever tools I have.
- 4:24Okay.
- 4:25And how long has that journey taken you so far?
- 4:30I've been at MAP for about four and a half years now.
- 4:33And that transition has about taken about a year
- 4:39or something for me to happen
- 4:42because I have been the product matter expert
- 4:45for subject matter expert for the mobile push
- 4:48and channels domain, SDKs domain.
- 4:52And I was contributing heavily,
- 4:53so it made sense for the business
- 4:55to actually transition me into the role.
- 4:59And here's another learning thing.
- 5:02I learned in this experience
- 5:03that it is not about the title that is given to you,
- 5:08and then you start working on it.
- 5:09It is rather the ownership and the contribution
- 5:13that you can do from day one
- 5:15without even having the title.
- 5:16And then title would eventually come to you
- 5:19when it makes sense to the business.
- 5:21Okay.
- 5:22So that sounds great.
- 5:23That's a great lesson.
- 5:24I mean, it's almost maybe it sort of dovetails
- 5:27into my normal sort of next question,
- 5:29which is, you know,
- 5:30if you were starting your tech career again tomorrow,
- 5:32knowing what you know now
- 5:34with all the changes in tech, right?
- 5:36The transitions in tech.
- 5:37I mean, I've seen loads of different technology changes.
- 5:41You know, what is the one thing
- 5:42that you would do differently,
- 5:44or the one thing you would tell your younger self,
- 5:47you know, as a bit of a lesson, if you will,
- 5:50to get that quicker?
- 5:53I'm glad that you asked this question
- 5:54because I wish somebody could have told me
- 5:57when I was starting my career
- 5:59that don't be afraid of talking to the customers.
- 6:02Being customer-centric is actually very, very good.
- 6:06In this world of SaaS products,
- 6:10when you try to look from an engineering perspective,
- 6:13I come from that background and that mindset.
- 6:16Let me build something.
- 6:18And this is how a customer would use it.
- 6:20But is that even solving the problem?
- 6:23So being customer-centric actually helped me.
- 6:27And if only if somebody told me this,
- 6:30so in my early careers,
- 6:34it wasn't easy for me to talk to the customers.
- 6:37This is something that I would have avoided throughout.
- 6:40So, yeah.
- 6:42That is a great, great lesson.
- 6:44I think it's one of those things
- 6:47that people may feel a bit unnatural about,
- 6:49but actually you realize that customers
- 6:52are just people like you and I, right?
- 6:54And if you just have a conversation,
- 6:56you will actually learn quite a bit from them.
- 6:59Exactly.
- 7:00And do you have an example where you've spoken
- 7:04to a customer that has directly influenced
- 7:08the direction or the developments
- 7:10that Map Digital has taken?
- 7:13Sure.
- 7:14I mean, of Map, I have plenty of examples.
- 7:17But on the side, what I also do is,
- 7:20in order to stay relevant in the industry
- 7:22and to not lose the touch of technology,
- 7:25yes, consumer, but also the customer,
- 7:27to not lose the touch of technology,
- 7:29yes, consumer-centric thing is important,
- 7:32but at the same time, being relevant
- 7:34and aware about what's going on in the technology
- 7:36is important.
- 7:37I work on my own projects via WorkRush,
- 7:41which is where I develop products for my own clients.
- 7:46I've got an opportunity to work with Ministry of Defense UK,
- 7:49and I've worked with University of Pyrnmouth,
- 7:52University of Bristol in this whole journey.
- 7:55And that actually taught me the thought leadership,
- 7:59looking at the problems from the customer's perspective.
- 8:04And when you talk to these people,
- 8:05when you talk to the business owners
- 8:06who actually got an idea
- 8:08and who are trying to build something very unique
- 8:11from their perspective,
- 8:13understanding what is the uniqueness,
- 8:14what is the value that they are trying to add to the world?
- 8:19What is the value that they are trying to get
- 8:21from this conversation?
- 8:23And how can you play a part in their journey of success?
- 8:28That, I believe, has really encouraged me
- 8:32to explore more in technological side
- 8:35and deliver up to the expectations,
- 8:38not just deliver, but based on the experience,
- 8:40I was also able to suggest them best possible ways
- 8:43and future-proofing ways.
- 8:45I work a lot on AI,
- 8:47to an extent that I also released
- 8:49one of the open-source standards.
- 8:51I created one of the standards
- 8:53called the Emotional Governance for AI.
- 8:58And you can find it on aiceels.org,
- 9:01which is the open standard that I wrote myself
- 9:04or the governance part of AI.
- 9:07And that tells me,
- 9:09it is important to look at a business
- 9:11from all these aspects.
- 9:12One is the commercial, then the technological,
- 9:15then is the future,
- 9:16which most of the people miss to think,
- 9:20or they think about it at a very, very late stage.
- 9:23How is the future look like?
- 9:26And is what we are building future-ready, future-proof?
- 9:31And what's coming up next?
- 9:32Are we going to be in that race
- 9:35or are we obsolete from day one?
- 9:40So that sounds very exciting.
- 9:43And so you have your own consulting business on the sides
- 9:48with a number of clients in the UK,
- 9:52including serving public sector,
- 9:55let's say public sector, government sector as well.
- 9:58That's it.
- 9:59And I'd love to know more about that as well
- 10:03on maybe a slightly separate conversation,
- 10:05but as, well, actually, let me ask you,
- 10:12when did you start that business?
- 10:14Because clearly you've been working in your day job
- 10:18at Map and other places,
- 10:20but so how did you get this sort of started along the side
- 10:23and how's it been going?
- 10:25So I had been a developer by heart,
- 10:28but development wasn't a part of my job
- 10:31when I started working.
- 10:33Mac tech is not a development heavy thing.
- 10:35And I didn't want to leave that one.
- 10:37So back in 2021 is when I started working on WorkRush.
- 10:43And I work alongside a couple of my other friends
- 10:48and we develop solutions.
- 10:49We get on calls with the customers
- 10:53and we understand the problem
- 10:55and provide a solution for them.
- 10:58Sometimes it is technological.
- 11:00Sometimes it's AI driven.
- 11:01Sometimes it's just,
- 11:03they wanted to have a second opinion
- 11:04from a thought leadership,
- 11:06from a technical perspective
- 11:09of how their business can work.
- 11:11In this whole journey,
- 11:12I've also worked with a couple of venture capitalists
- 11:16building their tech products
- 11:18and giving them a building a portal for them
- 11:21to assess other ideas, business ideas,
- 11:24and taking an opportunity for them to present
- 11:28to the entrepreneurs or the upcoming entrepreneurs.
- 11:34That's amazing.
- 11:35That's very exciting.
- 11:36So with that made a context,
- 11:39and what does the next chapter of your career look like?
- 11:43Right?
- 11:44I mean, where do you see yourself going
- 11:45in the next two, three, four,
- 11:48or maybe longer period of time?
- 11:50What does it look like for you?
- 11:51What's your ambition?
- 11:53Well, being in the tech industry,
- 11:56if I say job, then I want to get into Google
- 12:00as a product manager at Google.
- 12:02That is what I have always hoped for.
- 12:05But keeping ambitions on the side
- 12:08and keeping pure passion,
- 12:10I want to build something
- 12:12which is very sustainable for the future.
- 12:14So having my own product,
- 12:16which is sustainable, usable,
- 12:19should start with the open source,
- 12:20and I'm working on a couple of ideas myself.
- 12:24And with this whole changing world,
- 12:27I want to just create a harness
- 12:29so AI doesn't take over, let's say, in other words.
- 12:34And there is a governance layer everywhere.
- 12:36And I made the foundation with this AI.
- 12:38It feels so urgent, the open source standard.
- 12:40And eventually, I am planning to also launch
- 12:44a couple of other open source standards.
- 12:46So there are checks and measures
- 12:49for the future softwares that are going to be built,
- 12:52both from the security
- 12:53as well as from the development perspective.
- 12:55There was one research from Anthropic
- 12:58that said with the cloud code,
- 13:02developers are getting more,
- 13:06how should I put it, incognito.
- 13:08I don't want to use the term lazy,
- 13:11but they are losing their cognitive capability
- 13:17of understanding the code, reading through the code,
- 13:20because the coding AI agents are becoming so good, right?
- 13:24But there has to be a human element to it,
- 13:26not just approve and reject when you're whiteboarding.
- 13:29Not that everybody can code.
- 13:31There are also loopholes that AI knows well,
- 13:33because if a product is AI coded,
- 13:36you know, AI also knows the loopholes of those things.
- 13:42So just creating a guardrail, a harness,
- 13:45so the future looks pretty secure,
- 13:47stable, and sustainable for the humankind.
- 13:51This is what I'm trying to do.
- 13:52And so for you, it's about sustainable AI
- 13:57and responsible AI?
- 14:01Is that sort of what I'm gathering here?
- 14:04Yes, it's a sustainable, secure software building
- 14:09that I am building around,
- 14:12that I'm currently building the foundations of.
- 14:15So a secure software building, a SaaS product,
- 14:18because, you know, they are calling it SaaS Eclipse,
- 14:22because AI can build so many things and so quickly.
- 14:25There are so many softwares out there now.
- 14:28And there is a noise in the market,
- 14:31but with that noise also comes a lot of security challenges,
- 14:34security issues.
- 14:35Most of these products are white-coded.
- 14:38AI has coded them.
- 14:39Leaving all the loopholes that you can imagine,
- 14:43a regular, natural senior software engineer
- 14:46would have caught,
- 14:47or the compliance would have caught.
- 14:51But since these SaaS products are mostly started
- 14:55by a team of one or two non-engineers,
- 14:59just, you know, product mindset people,
- 15:02and they've white-coded the solutions,
- 15:04and it is out there.
- 15:05You know, people are putting their personal details onto it
- 15:09just to try the product out,
- 15:11and turns out tomorrow the product got hacked
- 15:14because the basic authentication hashing
- 15:16wasn't put in place.
- 15:18AI wasn't informed about it.
- 15:21There were no predefined checks.
- 15:24There are no evals put in place.
- 15:26So I'm just trying to build a community of people
- 15:30and laying the right foundation of secure
- 15:32and safe SaaS product building,
- 15:35so software building,
- 15:38using the AI,
- 15:39because AI is inevitable.
- 15:41Yeah, it is inevitable.
- 15:44And I guess sometimes if you don't know
- 15:46what to ask AI to do,
- 15:47then the output will not be as good as it could be, right?
- 15:54So my final question then,
- 15:55as we're on this sort of thought process,
- 15:57why are you so passionate about this part of AI
- 16:01and creating standards,
- 16:05open-source standards and things like that?
- 16:08From the early days when AI came in,
- 16:11at MAP I had been trying to build bots
- 16:14and meaningful operation of AI apart from bots.
- 16:21So it is far more than just chat-based interface
- 16:26or a voice-based interface.
- 16:28I want AI to actually evolve
- 16:30for a betterment in efficiency increase
- 16:35when AI does things for you,
- 16:36which client-like cloud will work,
- 16:41does that for you, right?
- 16:43But then are we then restricted
- 16:45to a monopolized environment of cloud code
- 16:48or open-AI's chat GPT?
- 16:50What else, what are the alternate options
- 16:52to be having the market?
- 16:53Using the same capabilities
- 16:55to the same level of power that these models bring.
- 17:01This is where my passion grows,
- 17:05that it should be decentralized and not monopolized,
- 17:09because capitalism is winning the world,
- 17:12but we can always put measures in place to check that.
- 17:18And that's where the whole motivation comes in.
- 17:21Love that.
- 17:22Well, what's the best way for people
- 17:25to get in touch with you, Deep?
- 17:26Well, I am available on LinkedIn.
- 17:30debunker-mathur01 is where LinkedIn puts me on.
- 17:36And my email ID is also debunker-mathur01
- 17:40at gmail.com is my email ID.
- 17:42So love to hear it from anyone
- 17:44who is interested in getting in touch.
- 17:46That's good.
- 17:47Well, Deep, thank you so much.
- 17:50Thank you so much.
- 17:50Amazing, it's recommended everyone have a look
- 17:53at what you have to offer here.
- 17:55And I appreciate you being on the podcast today.
- 17:58Really appreciate it.
- 17:59As a reminder to everyone listening,
- 18:01this show is sponsored by the Consulting Launch Method,
- 18:04where I help experienced professionals
- 18:06build their own consulting business alongside their day job
- 18:09without spending years figuring out the business side.
- 18:12We've helped clients go from first idea to first clients
- 18:17to within 90 days and into six figures within a year.
- 18:20To learn more, connect with me on LinkedIn
- 18:21and send me a message.
- 18:23Thank you, Deep.
- 18:24Thanks, Martin.
- 18:26Until the next episode, take care.
- 18:28Thanks, bye-bye.