Welcome to Pulse of Tech by C. Pune's leading technology podcast from C the city's premier platform for tech organizations delivering cutting edge discussions and insights. We often believe that the path to a tech stardom either starts from a computer lab or a coding boot camp. But for our guest today, it started with a childhood habit of identifying patterns. Today we are talking to Manish Dada who's the chief mentor for Isertis and the co-founder of his ninth venture Humanronic. Let's welcome Monish today.
Hi Manish, how are you?
I'm good. I'm good Parad.
Yeah, I mean it's a pleasure having you first of all on this episode of Pulse of Tech by SE. So welcome to the episode. Thank you so much. Thank you so much for having me.
So uh Manish, I want to start with a with the first question around uh your like history where um you returned to India.
Makes me feel ancient when you say history.
Not like that. Not like that.
I'm kidding. I'm kidding.
Yeah. You returned to India like mid PhD to fulfill a promise to your mother. Looking back, do you think it was like an um logical move or or an emotional move back then?
So, it's very interesting. I was loving my PhD. I was loving my masters. At some point, I got this moment which said, you know, where is this PhD leading to? It's too much work, right? And not enough doing uh kind of. So, uh and then of course, I promised my mom that I would come back as soon as I finished my education and not settle back in the US. But you know how moms and sons and promises go, right? But the good thing was I think I thought I would actually be more comfortable coming back to India than I would be living in the US. Though I loved living in the US, it didn't feel home even after three and a half years.
And remember those times in the u like late '80s were very different.
Right. So you couldn't call. You didn't you didn't have Skype. You didn't have anything. Calling was like I remember once in a week kind of a thing.
Oh my goodness. No. Once a month because we had to save money to make a call.
Yeah. Trunk call kind of a thing.
Exactly. Exactly. So we had to be careful with the money. We had to be careful with calling back. So 3 and a half years was quite a long time to be away uh from your family and we are a very close family. Even today we're extremely close. So I think there was a lot of emotion in there, less logic. Um, so it also boils down to serendipity, right? You take some decisions and you figure out, hey, this was wrong because of that or this was right because of that. This was one of these things. There was no there was more emotion than logic for sure.
Okay. And now reflecting back on it uh do you I mean obviously it has turned out well. So whatever I did, I'm sure it would have turned out well, right? So is like the reflection of it is what how do you want to reflect on it now about
That's a good question I think what I have seen uh in these situations is um you you take decisions and you have data at the moment and you take those decisions once you have taken them you move on you never go back because you did take the right decision at the right time whether it turned out to be wrong or right only time can tell but you have to take a decision if you don't it's like you know writing code with no bugs. You can't do that. The best way to do it is not to write any code.
Right. So you have to take decisions. Once you've taken decisions, you don't look back. You don't regret. Um and I've been I've been fortunate enough I'm very grateful that the decision I took actually worked out very well. You know those were times of the brain drain. Everybody every other week there would be a big article in the news saying hey you know the government is spending so much money on every student in IITs and everybody is going out out and I said hey I'm coming back you know it's do you want to fund me on this one
exactly
Want to take us back to that the moment so what what we have read is that you were a document cer for a boiler factory right and what is the exact question that that the executive asked you about forran that changed your life choices.
So great question by the way and great research uh by you and your team. I think it was very interesting. Uh I was in this first ever kind of um you know bachelor's program which was called a sandwich course. Okay. So you spent two semesters of your four years in a factory working um at a company on the shop floor. I was a mechanical engineer. So that's how I got into Thermax and I was not actually a document courier. It was even much more worse than that. I was a welder. So I was on the shop floor welding. Right. Okay. And then um my boss or the head of department actually not even my direct boss he said Manish you know I heard you know some programming somebody told me that you know some forran programming.
Mhm.
And I said yes and then he said what are you doing on the shop floor why don't you help me uh design some chimney design software.
Okay.
And I said what is you know what is it? And he said you know we you know for every boiler needs a chimney and the chimney design we take couple of weeks to do that.
Mhm. M
And it's fairly straightforward you know there are formula there's wind speeds there's height of the chimney there's sections that are required so he said if you know forran why don't you uh do this and I said I would love to but uh you know you have to give me some tools and he said no we have a sister company had a sister company which was an IT company at that time okay so he said why don't you go there I'll give you an intro uh and you should go work there and I said you know I I've also started learning C I would love to use C. And he said it's not about what you use. I just want to get the job done. If you have business value, you know, you don't care which tool gets the job done as long as it gets done right.
I mean, you have always described that the journey from 50 million to 250 million ARR is the hardest part of any startup journey. So um and and this has been like why why do you think that 90% of the companies fail in that part?
Yeah. No, that's a great question. Again, I think 0 to one is almost equally harder as 50 to 250 is, but there's a one fundamental difference in 50 to 250. What you as a founder or as founders, that's Samir with me as my co-founder uh do is fundamentally different than what you do in the 50 to 50 million journey. What do I mean by that? I mean you have to change your personality, your way of working, uh your learning ability and your delegation all at the same time in that journey. So by the time you get to the 250 million mark, I can vouch in my journey that I was a completely different person.
Yeah. Because you start scaling, right? You start scaling the problems that you solve are now completely different.
Okay. You're not at a level where you're operational anymore. You're more strategic. The decisions that you take not just affect a small group of people that you know well but hundreds of people sometimes thousands of people um that you don't know well and so you have to be really conscious of that fact and also that's when you scale not just the number of people you also scale uh the amount of work that you produce by when you say something or do something okay right this was this actually came to me late Samir understood this very well, it came late to me saying if I say something in a meeting, what are the repercussions of that?
Right. If I say that we will do this, what does it actually you know it's almost a domino effect. You end up having 5,000 actions in the company without even correct realizing that you started something. So thought leadership started playing a very important role from not just thought leadership. I mean you know at the 0 to 1 million also you need thought leadership. You have to figure out what is different your USP you know you have to think laterally think broad all of that happens I think it's it's a combination that happens at all at one time which makes it really really hard and that's where most people cannot cope up that's where most founders can fight that's where most founders start disagreeing because 50 million R you're actually a good company good company yeah so now you're thinking oh now I can you know either sell or I can do something I'm you know I'm already like successful but you're not there yet if your aspirations are much bigger so you start looking at things very differently and you start doing that it makes it very hard for you to change yourself because you think the world has to change because you know you've made something successful that never happens in real life right if you you have to change and your world around you has already changed.
And I think that realization was very interesting for me.
Okay. Okay. So, uh with that um I wanted to get into uh the part where you Isertis is your seventh startup, right? What is the fundamental problem which you are trying to solve uh in Isertis? I mean I have to ask like point blank on this.
No, absolutely. That's a great question. Uh and you'll be surprised to know many people are uh when Samir and I actually first talked about IT were not trying to solve a problem at all.
Um it was very interesting very interesting because both of us were coming out of acquisitions.
Um uh I blade logic had already gone public in 2007 um and on the NASDAQ and then 2008 we sold Blade Logic to BMC software.
Yes. 2009 we started Isertis. At that time Samir was also going through an acquisition. Okay. Um he was the CEO of Aztecsoft. He had sold Aztecsoft to mine tree and we both you know had a cup of coffee on Ferguson road and we said what should we do now? You know we can't sit at home you know um kalpa and sarikar wise wouldn't allow that. So it you know make sure that you do something. And we said what do we want to do? And we said both of us said you know I I have lots of ideas but nothing that I'm really passionate about yet. And Samir said but you know how do we actually think about what we have done so far. Um and as I said this was my seventh startup in Isertis. So you know I said I've done a lot of stuff. I learned you know a lot of distributed um systems you know I've built systems and somebody said you know I've also had a lot of fun but how do we actually think about building a consequential company out of India uh and we accidentally got into contract life cycle management and it was funny Samir called me one day and he said hey I ran into a friend of mine who is in Microsoft um he's a lawyer and um he said he just scrapped a project worth $18 million because he couldn't do what he wanted to do and I was telling him we we are in the Microsoft ecosystem we are building these ERP surround products and he said I don't understand that but can you build a CLM system for me so Manish what do you think and I said I know all about CLM you know in BMC after the acquisition had actually been involved very deeply in cloud life cycle management and I said I know about CLM and some he said how you know I I I don't know about it and I said you know that's what I did in BMC for for a year and he said what are you talking about? I said cloud life cycle management and he said no no no I mean contract life cycle management which is completely different and I said okay I don't know anything about that so okay that's how we started right okay and the fundamental problem we solve at isertis is to make sure that contracts which is a fundamental way to define a relationship between two entities is optimized okay right so you you want to make sure that you and optimize that relationship as much as you can.
Okay. Right. So that's what is fundamentally does. It solves that problem of how do you take a contract or a set of ideas or intents uh build it into a contract and make sure that in the real world it comes close to what you intended.
Okay.
So I have now two follow-up questions on this.
Absolutely. One is about because I mean it it is about Microsoft's relationship which you just spoke about and the other one is about the contract life cycle management as well. So my first question is about that uh your first uh major customer was Microsoft and now is a critical partner as well right so uh how do you navigate the balance between being natively on Azure and uh competing for the same seesuit attention uh as Microsoft there's no conflict of interest there or something Microsoft concentrates on productivity okay The whole co-pilot suite, office suite is really all about productivity. And of course, you know, there's the Azure business and so on so forth. But if you look at the co-pilot, we actually complement the copilot very well because we provide all the intelligence intelligence needed, the plugs needed for a contract management system to actually become better. And from there because we look at it not just from the management of the contract itself the intelligence behind what goes into the contract and was goes into the data that actually creates the contract and the peripheral metadata that is around the contract. Um so there is actually a very complimentary relationship and always has been uh with Microsoft and there is really no competition there. Right. See, I can't get away from not addressing the uh elephant in the room around agentic AI solutions and that's why I was talking to you about that contract life cycle management as well. I mean um so you have you've started advocating agentic AI uh where agents acts on act on contracts right um so I wanted to understand that uh uh if you are talking about an agentic solution and uh uh to manage the contract life cycle the agent can continuously reject a contract as well right uh how do you uh so this has been a question across the industry also that the human in the loop is very important for most of these things right because once the model starts hallucinating it'll keep doing the same so uh how do you keep that uh balance in in the solutions which you de
Yeah I think um you know with our Vera suite of AI solutions across the board the agents that we have one fundamental um you know kind of foundation of all that is responsible AI Right? How do I make sure that every action that is taken? How do I make sure that everything that the agent does goes through boundaries? It's almost like a contract, right? For an agent to work, you need boundaries. You need a construct which says, 'Hey, you cannot go beyond this uh on the left or the right or the front or the back. Uh so you want to make sure that those guardrails are in place.' Okay. So you put put those guardrails around that around that agent. Um with agentic AI I think very simply put when the agent determines or the human guardrails determine that an agent needs to go to a human it has to go to a human.
Um the way people do this today is a combination of prompt engineering and a combination of rules which says hey you don't have permissions but what it means is that your guardrails define where the human comes in right and the human comes in in two ways in my mind. One is the human in the loop. Okay, where the human comes in where there's an escalation where the guardrails need to be broken or the guardrails prevent you from taking action. Then the agent or the agentic orchestration needs to go to a human. But there's also something called a human on the loop.
Human looking at the loop and saying hey is this going as per the guardrails and is anything breaking and that is also as important as human in the loop. Okay. So the human on the loop actually look at what agents are doing, what their audits are, why they took decisions in this situation and not in that. Um you know what are the hallucination boundaries that you can figure out. Um lot of work has been gone has gone in over the years to make these agents very responsible and also making sure that they take the right actions and involve the right humans at the right time.
Okay. Okay. You you believe that heavily regulated industries like healthcare and finance will be the first big winners in AI. So uh most of the people think that these industries are too slow to innovate. Um why is the red tape actually a AI accelerator?
No. So very interesting. They're probably not going to be the first but they are going to be the big winners in AI in my mind. So think about it this what agents should not do right wherever there is you know broad acceptance like if you if you look at Instagram or tweet you know Twitter or whatever x Facebook LinkedIn etc you'll see a lot of posts you know this agent went crazy and deleted a database that agent went crazy and deleted a whole uh you know branch uh the thing is agents will histate that's the nature of LLMs till we fix that core problem on our way to agentic you know artificial general intelligence but till then you have to assume that agents will will hallucinate so you have to have guardrails now regulated industries have the best guard rails in the world right if you want to launch a new drug in the market as a pharmaceutical company you have to go through years of red tape to to to actually launch one molecule one molecule right so that compliance that structure that discipline in that process actually gives you some of the best guardrails there are in process. Once you have guardrails, the agents can do a lot more.
They already taken care of. They're already the human in the loop, the human on the loop, you know, where to interfere, where not to go, permissions, they're all already taken care of. Exactly. And very meticulously. Correct. If you look at a drug application or if you look at a like a clinical trial or if you look at from a financial services perspective you know how you calculate your uh debt mortgage calculator mortgage calculators even in the insurance industry if you look at you know how you calculate your premiums they're very detailed there's very detailed documentation that exists because the regulation is also that strong right okay exactly and that means the potential of the agentic you know kind of the whole world of agentic AI creating a lot of business value is very high in those industries. I have a fun question to ask you, right? Um I don't know if it is it is fun or not, but I'll still go ahead because chaos make it fun.
Yeah, let's make it fun. So, um chaos engineering is a word which is more around an engineering discipline, right? But uh but you introduce that in a very like a very interesting way in Isert it is in your teams. So, um first of all, I would want to understand so I have done my research. I know what it is, but I want the audience to know how do you introduce in it in your teams and just removing that that key person from your team. Doesn't it like just change everything in that team?
So, it's interesting you're referring to this experiment that you know I've been doing for a while. Um uh and I I won't call it an experiment now because it's a proven process. Yeah. Um, you have a well functioning team, you have a good leader, you take away the leader, you figure out what happens to the team.
Right.
And you also mentioned it that it's a little bit cruel, right? It sounds, you know, very cruel in terms of, you know, how the team feels. You know, suddenly, you know, I don't have a manager, I have a new manager, whatever that is. But what I found is very interesting. I think we as managers and I'm not talking leaders here specifically managers uh we as managers actually do not fully understand the potential of our teams. We always underestimate them because that's how we are wired usually.
Right.
And I'm making an average statement, not a very specific statement but most of us right and you know it includes me as well. Right? We un underestimate our teams. We always think hey a good leader will actually make a better team which is true. But what happens is if you have a very good leader then the leadership qualities in a team are repressed. So chaos engineering in this context actually tells me that if I take away a very good leader and let the team figure out who takes lead.
I'll find the next leaders. Right. our problem and I said 50 to 250 million dollar journey is the hardest uh one of the hardest journeys in the world in the life of a uh of a company I think this is one of the reasons this is one of the learnings I had that you you cannot just hire leadersh you have to nurture them
Right and if you're nurturing them you have to give them a stage and the only way to give them a stage is to force them to lead Y and this is a classic way of making a whole team understand hey figure out who wants to step up uh who wants to take that responsibility who is ready or who feels and is confidence that is ready and then support them from outside so actually it's not cruelty right
Yeah and of course when you take away the when you explain it in that way it doesn't look like but I mean it feels like
Yeah I mean I um I tend to agree with you completely on I've worked or built teams which are u which are like uh huge teams but the most important thing which I have also learned over all these years is that that rather than making somebody the leader of the team let the team select a leader out of it is always a better way of or not select actually let let some person take up that responsibility. responsibility has always proven much valuable than just putting a leader there. Actually I wanted to um uh talk about a sensitive topic here like uh the passing away of uh your co-founder Samir Buddha's uh earlier this year right must be a big impact to the company. So uh how has his legacy changed the way uh you view your forte values which both of you built together right?
So if you are okay to talk about it.
No, absolutely. I think whenever I think about Samir brings a big smile on my face. Uh we had so many lovely moments and so lovely memories. I think it's just I'm so grateful that I had the time that I had um with him and forte was very near and dear to our hearts when we came up with this whole construct of what our company's values should be. Um fairness, openness, respect, teamwork and execution. That's what forte stands for. Um we were thinking these should not be values on a wall. We were thinking hey if if I have to live them then and I expect other people to live them then they have to be real.
They have to be pragmatic
They have to be not just practiced but they have to be refined over a period of time. I'll give you an example. Um this was Sam's favorite example. You know used to say when you go to the US Manish when you come to the US he was he lived in the US for many many years um and at that time we we all carried visiting cards right business cards and it was very normal when you walked into a meeting to pass on business cards to people and usually if the person was on the other side of the table you would just throw the card so that it would reach him or or her across the table and he said would you do that in Japan oh no I said no absolutely Not it's like it's disrespect.
