Beyond the Simplistic Narrative that AI will Replace Software with Mahesh Rajasekharan

Speaker A: Uh, I believe AI will reshape SaaS, but not in the way most people think. The narrative that AI will replace software is way too simplistic. There will absolutely be winners and losers. But the dividing line isn't whether a company has AI. It's really about whether a company owns the context required to create outcomes. For example, if a software application performs relatively simple workflows, in that case, the frontier models will increasingly be able to replicate those outcomes directly, and so those businesses will face significant pressure.

Speaker B: This is SaaS scaled the podcast where data meets action with host Arman Shrakhi. Each week, Arman will be sitting down with CEOs and industry leaders from the technology sector, giving you the insight to innovate without reinventing the wheel. They'll discuss challenges, best practices, and how to identify the right metrics. So if you want to get to market faster and in a way that matters, then subscribe and join us every week as we discuss SaaS Scale. This episode is brought to you by Curve, A the modern no code analytics solution. The tools you need to take action with your data on a platform built for maximum scalability, security and cost efficiencies. If you're ready to reduce complexity and dramatically lower costs, then contact us today@crave.com that's Q R-V-E-Y.com.

Speaker C: Hello, welcome to another episode of SAS Scale. I have a fantastic guest this time and Mahesh is joining me from Clio. A fantastic company and also very impressive background of Moish himself that we will get into it. But before you know, we get into some questions I have and learning more about the company and Moesh and what he has done with the company. Uh, I will let him introduce himself and the company a little bit and uh, we get the context so we can stop.

Speaker A: Thank you Arman for um, inviting me to your podcast. Uh, my name is Mahesh Rajasekaran. I'm um, the president and CEO of Clio, the global leader in supply chain orchestration solutions. Uh, we have an innovative platform called Clio Integration Cloud and we have supply chain orchestration capabilities that enables us to drive the digital ecosystems of thousands of customers and partners. We do that by providing highly flexible industry focused solutions that really bridges the gap between supply chain planning and supply chain execution to deliver better real time decision making which is so critical in today's constantly changing business environment. Um, my background spans more than two decades across business strategy, deep work in supply chain planning and execution, um, as well as product innovation building, uh, go to market across sales Marketing and operations. Um, from a CLIO perspective, I would define CLIO as a supply chain orchestration platform that helps organizations automate, connect and gain end to end visibility into the trading partner ecosystems. It enables businesses to streamline EDI and API workflows, accelerate partner onboarding and proactively manage data flows to improve operational efficiency and reduce manual errors.

Speaker C: I'd like to ask you a question before starting the main questions, but the clarifying question you mentioned, orchestration platform. For me and the rest of audience who may not be super kind of clear on this, how do you differentiate uh, orchestration platform versus integration platform?

Speaker A: Fantastic question. Um, orchestration is about looking at core business processes like order to cash, procure, to pay, lotion to invoice, and make sure there is a, uh, complete choreography of the business processes across systems, across business transactions and across Personas in a business such as a customer success, uh, rep, uh, fulfillment specialist, a procurement specialist, uh, logistics coordinator. Integration platform on the other hand, makes sure that the integration signals, such as an order, a purchase order, um, advance ship notice, an invoice, a point of sale, an inventory snapshot is accurately reflected and connected between the participants in a supply chain such as the customers, the company, their suppliers and the logistics providers. So integration gets all the transactions connected, all the systems and partners connected and the transactions flowing. And orchestration takes the next level and really choreographs the flow of information across business processes so that there is intelligence and proactive decision making. So one feeds the other. More integration signals you get, the better you can orchestrate.

Speaker C: Fantastic. So essentially uh, I would say the orchestration platform is the next generation and more comprehensive platform compared to the integration platform that you put the integration in place but then you complete it, you go to the next phase and when you have the orchestration platform, uh, you have a more real time synchronization. Everything works in a well orchestrated way. Real time essentially.

Speaker A: That is exactly right, Armand. Um, because if you look at supply chains there is so much of information flowing and the decision making has to keep up with changes happening. Whether it's in terms of tariffs, in terms of uh, supplier, uh, challenges, in terms of logistics, uh, capacity constraints, the planning application, what we call the supply chain planning applications for forecasting and planning and the fulfillment, they have all the context but they just don't have the granular data to do real time decision making. And most of the execution platforms are just focused on short term, sort of point to point execution. They really don't have the context. And so orchestration is about bringing Context to real time execution. And that's why the next level of value creation for the enterprise so that companies can capture more revenue, can reduce costs and create not only efficiency but better customer relationships which improves their supply chain performance and business performance.

Speaker C: Okay, that's ah, a good segue to the next question. If you draw the line and uh, say we are in the AI age now, B4AI and post AI, how do you see the impact of AI in what you guys do and what are the things that you think will change drastically and solve the things that probably would not change anyway if AI is there or not? Because those are still the things that you know, have to be done, uh, in a way that we have done always we have done it and you know, we will do it the same way, probably maybe uh, a little bit faster, but has to be done anyway. Outside AI.

Speaker A: Uh, first of all AI is so fascinating. It's uh, exponentially growing and it's probably one of the most seminal time in human history that we are living. So it's really exciting. Uh, I believe AI will reshape SaaS, but not in the way most people think. Right? The narrative that AI will replace software is way too simplistic. There will absolutely be winners and losers. But the dividing line is into whether a company has AI. It's really about whether a company owns the context required to create outcomes. For example, if a software application performs relatively simple workflows in that case the frontier models will increasingly be able to replicate those outcomes directly and so those businesses will face significant pressure. On the other hand, there will be winners, right? Where SaaS companies continue to win is where outcomes require deep domain expertise, industry specific knowledge, uh, such as in food and beverage logistics, CPG, automotive etc. Access to appropriate data, which is not the first party customer data, but data in your platform. How do you look at uh, correlations between supply chain entities, how we can normalize the data, complex workflows that span across multiple functions and systems governance. Trust, trust is extremely important for businesses. So the way I view this is AI will and may generate the intelligence, but software provides a context, controls and the execution layer. So the future belongs to those platforms that combine AI with agentic workflows and not just AI in isolation. So uh, I think Armand, it really leads to a second misconception in my mind that hey, software development becomes so trivial, uh, in an AI world. The reality is absolutely the code generation is becoming easier. You can generate a lot of code and people actually were able to generate a lot of code even before AI. But the key thing is that generating code is only the beginning of building great software. The harder and more valuable work is everything that comes after, which is testing, validation, observability, uh, scalability, reliability, security, governance, trust, um, ongoing maintenance. A lot of times when ah, a uh, bug report is filed. How quickly can you make a change to the platform? How quickly can it be uh, in production can you add value, adding capabilities and add additional modules on the platform? So I view that enterprises don't buy code, enterprises buy outcomes, trust and accountability. And so these responsibilities don't disappear with AI, they actually become even more important. And the exciting part is that AI dramatically expands what great software companies can build. And we're observing it firsthand At Clio we're able to do a lot more than we even imagined two years ago. We can take a five year roadmap and crunch it in 18 months. For the first time, teams can move at a speed that is previously unimaginable. Product concepts can be developed, iterated, tested, refined and deployed in a fraction of the time. So I believe the best companies won't simply build the same products faster, they will build products that were previously impossible because economics, the talent requirements or development timelines were just prohibitive. And Finally, I think SaaS companies now need to start thinking about their software differently. In the past, before AI, historically software is built for human users. It's really um, human led, technology assisted work. Now we're transitioning into a world of human supervised technology driven operations. So the software must be built for AI agents. So I believe in the future agents will participate in decision making loops across every business process in the enterprise. The agents will discover information, they will evaluate options, um, and they will make recommendations and initiate actions. So it's important for the software platforms to expose a rich context, trusted data which is properly governed and secure. And you have agent friendly interfaces that will become part of those workflows. So I think platforms that remain optimized only for humans, ah, and human interaction will risk becoming marginalized. And those platforms that are uh, agent first and are agent friendly will start to get massive agency in the era of AI. So the next generation of market leaders won't be the companies that simply add AI features. They will be those companies that combine proprietary data, domain expertise, industry knowledge and agentic execution into systems that consistently deliver business outcomes.

Speaker C: Um, this is really about uh, adding AI to software, but in a way that is pragmatic, not just innovative. And it's based on customers and users need. And I think that uh, brings the kind of power of execution to the table because if you really don't execute it well by just you know, going too fast or going too slow or not doing enough to really make sense, you know, the way you add AI into your product so customers really don't see it in a kind of valuable way while you're adding cost by really integrating and adding AI. I think those are the optimizations that probably software companies have to go through during the next few years to really make the right decision, go at the right pace, you know, bring the right capabilities that users care to users and then uh, understanding which one is a sales asset and which one is a demo asset versus which one is ready for production, you know, those kind of things that all the decisions, challenging decisions, difficult decisions that software leaders need to make. Moving forward from your perspective when it comes to really uh, you know, going forward with AI in a way that you think is the best way and you really, the idealistic way that I want this to happen as soon as possible, how do you see the kind of challenges to be the way, I mean you experienced it? You, you have a very solid foundation from experience perspective in software world, not so many people can claim to be as experienced as you are. You have been, you know, uh, working with many investors sitting on different boards. So when you tell us something I personally will take it very seriously is coming from someone with so much experience. And I wanted to know when you look at AI and you look at software that you know very well, what are the top three things that you come to your mind and you say these are the things I need to be very careful about the challenges that I will see and I need to really make sure they are executed well when we do, otherwise they may become a problem. In adding AI to our product, the

Speaker A: most important thing which you actually highlighted pretty well is really solving problems for customers. In other words, you create outcomes uh, that are solving pressing and big business challenges. I think that's important. And uh, so in other words it's about engineering and not coding. A lot of people confuse AI can generate code and create a lot of code. But what you need is you need great engineers that can engineer solutions to business problems. So that's most important. Always have uh, problems in mind and you engineer outcomes and leverage AI to deliver better solutions fast. The second thing is there is a significant change management required, uh, not only within software companies as they're becoming agentic, but also with the end customers because the historic metrics have to change, the historic org structures have to Change. In the world of agents, the most important thing is to have agents that act, not just have agents that merely suggest, right? The agent that merely suggests are nothing but business intelligence showing up and telling you different things which most people actually will ignore. Because when you're getting a lot of um, alerts and signals and notifications, people don't look at them, right. So what is really important is to understand what the customer end outcome is, look at all of the signals across the enterprise, rank order and filter it and communicate to the human in the loop the three things that needs to be addressed in the next 30 minutes. For example, in supply chain, this particular performance issue in terms of uh, meeting an on time and full metric, which is a very key metric for retailers is to reroute a shipment or be able to fulfill it from a different warehouse paying an expedited charge with a logistics company. So those decisions in a timely prioritized manner becomes important and the organizations, uh, the companies actually should be able to redefine the metrics to be part of a agentic world where there's a combination of agents and humans in uh, every business process. And the third thing I think which is really important is you cannot just sell agentic workflows and agents and expect customers to just go and deploy them. You have to actually spend time understanding the business, understanding the business flow map, the as is processes, right? Like a swim lane, you know which Persona is doing which step, what they're interacting with, what phone calls they're making. For example, you know, somebody in the retail supply chain may take a phone call from Amazon or Walmart because um, the retailer wants to know where the shipment is. Now they will be making some phone calls to their warehouse. They may be making phone calls to their transportation desk. They might be logging into ServiceNow to look at certain um, certain process step completions. All of those would understand, identify the bottlenecks and come back with a new way of doing things. And this is where we need forward deployed engineers. We call it an AI SWAT team within clio where an AI SWAT team which is engineers, product people, solution, people who can understand the business processes of the customers, agree with them on the next step, have the agents embedded in the workflow and execute which means leadership both in the software companies to have a value centric uh, and an AI SWAT team approach and from customers in embracing AI into their businesses. I think those are the challenges but also I think phenomenal opportunities in the agentic era for companies to significantly improve the supply chain performance which leads to significantly Improved business performance.

Speaker C: Now it's time for a very random question. You have been probably traveling a lot. You have seen many cities in the world. If you wanted to tell me about one of those and say Armen, if you have two weeks, just go to this particular place. Uh, I had a great time there. It's a fantastic place. Either small or large, it doesn't matter. But what would be that city that you would say this is a great place to visit? For sure.

Speaker A: Um, I, I travel a lot globally. Um, one of this, one of the cities I, I went there many, many years ago and which still has a special place is Amsterdam. The reason is, um, the whole Benelux region, the Belgium, Netherlands, Luxembourg. The region, uh, to me is a gateway to Europe. Ah, A lot of the fast moving consumer goods, a lot of the supply chain, a lot of the logistics happens there. The people are very friendly, the people are very open, uh, and very collaborative. Um, which is not to say that it's not the same elsewhere. I actually uh, find traveling and meeting people and understanding different cultures extremely important, um, for a CEO to really understand not only the companies and the customers we interact with, but also learn about their customers, their suppliers. But I would say, Armand, that Amsterdam has been a special city. Uh, great, uh, great history, great culture, but also great people to work with.

Speaker C: So that's good, Mahesh, because you voted for a city that it seems like now be the front runner. I have asked this question now probably I don't know, I, I don't ask that question. I was not asking this question in the early episodes, but I'm starting to ask that question more often probably. I have asked it 50 times now and uh, I think Armstrong is the front runner now. So you are voting to a city that probably has received three other voting so far, if I'm correct. So that's great. And uh, now thinking about AI, about integration of different systems, do you see that using AI now integrations might become much easier than it ever has been. So does it have a positive income outcome, um, when it comes to integration? Because it has always been very challenging part of software that you write the software applications and then integrating them to work with each other very well takes more time than writing each of those parts and applications, whatever they are. But now with AI Place, how do you see it? As someone who's very experienced in that

Speaker A: domain, it's absolutely, uh, an interesting time and a great opportunity. You're exactly right that integration is way harder than people think because uh, it's not just the Surface level, connection between the systems or ah, business processes. It's all of the organizational learning, all of the customization. Uh, most of the things never get fully documented. The power of AI and then the way we are designing our agent swabs is we go and understand all of the other integration code that is ever written. It could be package software, it could be, you know it um departments could have written their own code to integrate. Uh, they might have made some assumptions, they could have made some customizations. Now we have the most powerful uh capability with a set of agents to understand all of that so that we can come up with a much cleaner way of integrating systems so they can respond to real time. Um, as an example we talked about why orchestration is a holy grail, uh because it's making real time decisions. But to execute real time decisions you need to integrate, you need to connect the ERP system, the transportation management system, the warehouse management system, sometimes your um, systems and engagement, uh, like a salesforce, um, like a servicenow to choreograph the process. And so the ability to have integration agility is critical for intelligence and orchestration speed.

Speaker C: Now question now outside the integration and software in a way, but it's still related to AI and software. If your cousin, your best friend, uh, your brother, someone very close to you came to you today and said I wanted to start a software company and this is 2026 with everything you know about AI software the way we are doing business, what would you recommend that person to start the company in the way that you would say this is a good domain, this is a good area to go and think about it. Because I think software applications or software companies created and when I say software in a very AI coded but this is really the kind of domain that you would suggest to your close friend to really go and think about it and consider that to start

Speaker A: there are definitely uh, few important domains. I think healthcare is very important. I think with AI, uh we can get to a world I believe in the next 15, 20 years of almost a deceaseless society. Because AI has analytical power to look at information at scale, uh, which no traditional drug discovery despite billions of investment can do, uh, to be able to not only do uh, drug discovery in terms of detailed disease progression but also actually build highly efficacious drugs. Healthcare I think is a very very critical area and I think uh, that's something I uh, deeply recommend. And the other area is very much close to my heart which is supply chain. Because if you look at what's happening today in AI, there is energy shortfall, there is need for chips, there is need for infrastructure, the need for compute. Uh, all of them requires the supply chains to be in place before you can get to the model layer, before you can even get to that application layer. Right. You need all of the other pieces to work flawlessly globally. And so there are so many aspects of supply chain which are not solved. In fact, uh, there is a report that Accenture put out, uh, last year which said just in the United States, uh, inefficiencies in supply chain are costing companies 1.6 trillion every year. And that 1.6 trillion with a T every year roughly equates to about 50%, uh, of the net income of companies. So imagine companies can double the net income by addressing supply chain. So I would recommend a friend of mine, uh, to absolutely look at all the things that can happen in supply chain in the era of AI and, um, um, really geopolitical tensions, which are really making supply chains, uh, strained. And so you need a lot of agility and intelligence.

Speaker C: Yeah, thanks for sharing the insight. It's very helpful. Um, and in the end, I would like to ask you for a book or a couple of books that you would recommend because you loved it.

Speaker A: Yeah, I read a lot of books. Uh, there are some books I just go back over and over and reread them. The very first book I recommend to anybody is a book called From Good to Great by Jim Collins. Good to Great is a fascinating book that I return to again and again. And I think of this book as a blueprint for building an exceptional company that can stand the test of time. And many of the principles from Good to Great have directly influenced how we are built. Clear. For example, one, uh, of the core principles in the book is getting the right people on the bus, getting the wrong people off the bus, and putting the right people in the right seats. It's all about talent. It's all about alignment. Uh, another principle is how do you build an organization that can consistently renew and improve itself? If you look at Clio, over the last 14 years, we renewed and transformed ourselves with a north Star of supply chain. But we went from moving data, uh, to integrating data, to enabling ecosystems in the cloud, to building a modern, uh, cloud network, to orchestrating global supply chains. That is very much about how do you renew yourself to solve the most challenging problems while you have a very clear north star of what you want to do to make the world better? Uh, a very key idea from the book is this idea of fly deal. If you want to build a great business, you need a lot of growth lovers that help your flywheel to keep cranking. The flywheel analogy is great because initially it takes a lot of people to crank the flywheel, right? You push hard and you push hard. And some, somewhere between the 7th and 8th, uh, cranking it becomes easier. And then because of gravity, the flywheel just compounds and it just powers its way through growing markets in a crushing competition. And the trick here is everything compounds, right? Uh, innovation compounds, product compounds, your customer relationships compounds, trust compounds, talent compounds, your brand equity compounds. And so I find it especially valuable because the lessons from this book goes across industries. And I reread every year and even more often and present on the book periodically because ideas take on new meaning as our company, ah, Clear. Grows and my own experience changes, uh, especially in the era of AI and working with customers, uh, as we navigate all these geopolitical tensions and opportunities. Um, if I can give another book, Armand. Ah, I think the second book is, um, another favorite of mine. It's called the Hard Thing about Hard Things by Ben Horowitz from Andreessen Horowitz. Um, and this book is especially relevant for CEOs. It really focuses on the very difficult and often lonely decisions leaders and CEOs and management teams must make during periods of uncertainty. And this book really shows what it takes, the courage of conviction to navigate crises. How do you make tough calls? How do you grind it out? How do you continue building an extraordinary company under stress, under crisis? And those lessons are particularly relevant during major inflection points, be it cloud and now the current transformation around AI. It really makes it extremely, uh, meaningful. Um, and there is one other book which I would also recommend. It's called who. Uh, it's a Method for Hiring by Jeff Smart and Randy Street. And why this book is so important is because it really reinforces how important is the discipline of hiring to building a successful company. Most people talk about hire and fire. I mean, that that's a bad idea. You really want to spend a lot of time hiring great people, right? A, uh, great strategy, a great vision will not matter without the right people to execute it. And hiring mistakes, especially at the leadership level, can have a significant impact on the entire organization. And it takes a long time to fix it. So ultimately, uh, the book talks about how companies win and retaining great people.

Speaker C: Thank you so much for sharing the books and, uh, I'm so glad that, uh, you had the time to join us. It was a pleasure speaking with you, Marsh.

Speaker A: It is such a great pleasure talking to you, Arman. And the questions are so pertinent, especially about, um, AI reshaping SaaS and what it takes for great companies and great software companies to build outstanding platforms.

Speaker B: Thank you for listening to SaaS scaled with Arman Eshragi. For show notes and any resources mentioned in today's episode, go to saasscale.com if you're enjoying our show, give us a five star review and share on LinkedIn. And be sure to subscribe for any updates on future episodes. Thanks for listening. This episode is brought to you by Curve A the modern no Code analytics solution. The tools you need to take action with your data on a platform built for maximum scalability, security and cost efficiencies. If you're ready to reduce complexity and dramatically lower costs, then contact us today@crave.com. that's Q R V E Y Com.