JD Dhaliwal: Powering the Next Wave of GenAI ISV Partnerships at AWS

Chip Rodgers  0:02  
Chip. Hey everyone, welcome back to another episode of Inside partnering. I'm Chip Rodgers, host and founder of inside partnering, and I am just so excited to be joined by Jagjit Dhaliwal and also known as JD, more commonly known as, JD, so. JD, welcome.

Jagjit Dhaliwal  0:26  
Thanks, Chip. I'm so excited to be here, and thanks for the opportunity. Looking forward to our conversation.

Chip Rodgers  0:32  
Yeah, yeah, so, so. JD, you know your your background is, is, really interesting and, and want to just give a little bit of an introduction to you, because I think you know your your your history starting out at TCS for, you know, number of years ago, for like seven years, and, you know, really in consulting, and then with Cognizant. And I think your shift actually. Then it was around then when you moved to the US and were focused in, you know, in US, US customers. And then you took an interesting, you know, tour of duty with LA County as deputy CIO, which is, which was a big role. And I'd love to hear about, hear about that, and how that, you know, your sort of learnings there in the public sector. And then the year and a half with UiPath, and now with AWS leading the you started really in the sort of Intelligent Automation, you know, with the UiPath automation anywhere, and you know, those technologies working with those partners and and then, more recently, now with Gen AI, which is, of course, the hot topic for everyone. A lot of partners, a lot of innovation, a lot of investment. So JD, you know, Thanks for Thanks for joining, and really excited to chat with you.

Jagjit Dhaliwal  2:01  
No, absolutely, as I said, honor is mine. Love to share any insights and experiences which I gather from all the large organizations. It's definitely a fascinating journey. Lasts about 2425 years, working across different life cycle of software, touching upon a different roles. I mean, it just gives you the perspective. It helps you understand what not to do. You learn a lot from that.

Chip Rodgers  2:30  
Isn't that the truth? You know? So let's talk about your let's start with your role at AWS, and you know, you're working with, with partners, with ISVs in particular, at, you know, AWS ISVs in the Gen AI space. And I think your role is, in a way, sort of two factors. It's, it's, in one sense, the, you know, working with partners that are Gen AI partners, and that's one sort of part of part of your role. And then the other part is taking all of the Gen AI capabilities and technologies that AWS has, bedrock and and you know, all of those, you know, Amazon Q, and all the capabilities that that Amazon, Amazon has, and encouraging and working on having the having partners. Could be Gen AI partners or other partners working with those technologies. So talk a little bit about, you know that where that is, what that role is all about, and how you're working with partners.

Jagjit Dhaliwal  3:39  
Yeah, no, it's a fascinating role. Took up this role last year. As you said, for last few years, have been focusing on Intelligent Automation, and the boundaries between automation and AI has been blurring for a while. I mean, if we go back, like 10 years, that's where robotic process automation came in, and then we pivoted to Intelligent Automation, which was nothing but adding the AI capabilities to RPA. And now with Intelligent Automation, we are adding the generative AI. That's where these two fields are coming together. And currently, as you rightly said, so my team is focused on those two pillars. One is we are primarily responsible for partner management of Gen AI ISPs in amber. So my team covers the full stack, from the large language models to ml ops platforms, vector databases, data, evals, testing, Agent tech platforms and IPAs platforms, which is enabling the agent capabilities, and then the AI infrastructure layer underneath itself, so it's a full spectrum of stack which my team supports. And then I roll up to Ammer ISV partner leader, so she covers all the ISVs. And if you. Think about Gen AI. It is not just about the Gen a ISVs. It is about, how do you leverage these Gen AI capabilities in enhancing the core capabilities and offerings of all other softwares? And that's what the second pillar of our responsibility is. How can we help and evangelize other ISVs to accelerate that journey, and it's their core capabilities and offering may remain same, but they can definitely enhance it using these capabilities. And AWS has our own offerings as well as we have a whole AWS ecosystem partner ecosystem, where we bring complementary capabilities so we expose those value propositions to other ISVs and help them accelerate that journey. And there are different sub categories within those scope, like business application layer, or you can think of industry vertical ISVs, or it could be the traditional data engineering ISVs, how they are actually leveraging and adding agentic capabilities to their data lifecycle. Security is another big element for agentic and Gen AI space. So how do you have security within Gen AI as well as, how do you use Gen AI within the security lifecycle? So that's where all these things are intermingling, and we are really helping owning the journey ISVs as well as we are evangelizing the other group. So never a dull moment and and you can imagine how things are moving at a fast pace here.

Chip Rodgers  6:35  
I'm sure it's, uh, these are everything's moving so quickly that it's hard to even keep up, you know, with the latest, I mean, everything's, you know, just innovation is happening so quickly. New ISVs, new startups, and, you know, just even trying to understand, okay, is this a real thing, you know? Is this really something that I should, you know, put some brain power into, to understand about, you know, and invest in, and how do you manage the pace of change? And you know, how those, how those, how those, the just how things are changing so quickly, with AWS partners,

Jagjit Dhaliwal  7:23  
yeah. I mean, if you compare these ISVs with the traditional ISVs, I think the key difference itself is the pace, and that's a two fold aspect to it. One is the pace at which they are innovating, when every other day we are hearing new models coming in and the new agent and capabilities coming in. And then other aspect is the revenue growth itself of these companies. I mean, these companies are setting a new benchmarks in a bar as compared to traditional SaaS. And I mean, I typically give example, if we look at Gartner Hype Cycle, typically from innovation trigger to productivity plateau. In the old age, we used to see that like it's a five year typical time frame when you see that emerging technology is kind of settling down into the into the core mainstream. But in case of Jenny, I am in this timeline has completely squeezed to probably like two to three year max. So we are seeing that kind of fast pace happening. And I was reading a report other day, very interesting report about the SAS benchmarks. So I mean, in the past, there used to be a benchmark for which we used to call it as a, t2, d3, which is basically a five year revenue growth cycle for SAS companies, the first two years they are hitting triple growth, and then next three years hitting a double growth. And there come the report actually talked about the new benchmark of q2, d3, basically first two year per ripple growth and then the triple growth in the next three years. So imagine that kind of ecosystem we are living in right now, where things are at a completely different pace. And end of a day, these companies are still startup, so still there, and the number of employees in these companies is very minimal. The revenue per employee is a completely different

Chip Rodgers  9:23  
I was going to say revenue per employee is also another thing,

Jagjit Dhaliwal  9:27  
completely different. So they may not have a large GTM team, they may not have a big partner team. So you cannot apply the same motions, and you cannot use the traditional mechanisms, which you used to do for SAS, and you have to come up with a tailored mechanism which is very agile and which has to be very responsive, partnership framework which we are activating for these so that's really set these partnerships aside from all of the traditional ISVs, which we used to manage in the past. Cool,

Chip Rodgers  10:01  
yeah, so it's, and, I mean, how do you, how are you managing that? Because you also have, AWS has a very robust organization that is just Nurturing startups, right? And, and so you, there's a, there's a whole separate team, but I'm sure you're working with them quite a bit as well. What's, what is, how does that sort of process work?

Jagjit Dhaliwal  10:27  
See, it is about, as you say, rightly said, AWS has a very mature partner ecosystem. We have so many programs and motions which are available to support a wide variety of partnerships, and it is our responsibility to narrow it down and figure out which motions are applicable and valuable for this set of partners. And then how can we tailor those motions specific to the needs of it? And as I said, like it has to be very agile and responsive. So we have to come up with the different mechanisms, even if we are using the same motions. How can we come up with a tailored approaches on it? Because I may not have a luxury to run the motion for six to 12 months and wait for the customer or partner feedback. We are pretty much trying those motions in the matter of weeks so and we have a reciprocative feedback loop from the partners. In the matter of weeks we are seeing that what's working not working. And adjust those programs. There are many pilots which we have tried in different incentive and funding programs like that. And then secondly, you cannot use the traditional manual ways of onboarding the partners, because there is an influx of agent companies which are coming in. So how can you automate and how can you make the onboarding process efficient? How can you make lot of those things self service enabled, and you can enable all these things at a larger scale to the volume which we are dealing with. And then how can you come up with a reusable assets, which you can reuse across the partners to really build that momentum, which you need to do that? And then goes back to their minimal team size. Lot of these companies are plg focus. That's what their GTM focus is. And so we have enabled lot of programs and motions aligned with that. I mean, if you look at even marketplace time, program is specifically targeting plg, and we are activating a self service customer abilities in that those

Chip Rodgers  12:41  
are the kind of trials. And all. Yeah, exactly, yeah. So

Jagjit Dhaliwal  12:45  
those are all the things which we focus on with the sole objective of, how can we make a Mac How can we maximize the impact with the minimal effort from partner? That's what they look for. So how do we do that? And how do we help them prioritize those things, because there is a whole buffet of different GTM motions there. So we have to help tailor it out. We have to narrow it down and see which motion will make sense for which layer of Gen AI stack, and then constantly keep on evaluating and evolving those daily basis.

Chip Rodgers  13:21  
Yeah. Yeah. I mean, I think you guys met Matt announced it was maybe three weeks two or three weeks ago, four weeks ago, the on the on marketplace, the agentic solutions. And it was, I think, when it was first announced, it was 900 and then quickly went to like, 1200 I'm not sure where it is now, but that's a high volume. And so I absolutely get your point that you can't do that without having automation and easy, you know, well understood or easy to follow workflows,

Jagjit Dhaliwal  13:59  
right? No, absolutely. And I mean agentic AI marketplace, that's exactly was our intent. Because 1200 is today. I'm pretty sure the number is higher today. It's increasing every day. And by reinvent this number gonna be way higher. So and the intent of introducing agentic AI marketplace is exactly to manage this influx of the volume. How can we make that experience seamless for customers as well as for our field sellers and the partner sellers, and that is why we came up with this categorization model, that how can we further sub categorize so that customers and field sellers can identify and find out the right solutions, what they are looking for. I mean, as you probably have seen, that there are five different categories of agentic solutions, what we have on marketplace, be it pre built agents, be it agent development, Agent tools, professional services, or the software solution which is leveraging agentic. Capabilities. So that's our first level of categorizations, which we have done. And then secondly, if you look at our other programs, like our competency programs, that's also helping the field sellers and partners to really look at for these type of solutions, what are the top partners which have already validated those stringent requirement of those competencies, we have those Gen AI competencies, which we have activated, which is helping partners to do that. And then other thing which we are enforcing is because we know that the volume gonna increase significantly is, how can we ensure that partners gives us a very detailed description and prescriptive documentation around the type of solution they are coming up, so that as we are using the NLP based search criteria across different mechanisms, so we'll be able to narrow down that things and make it easier for the field, sellers and the customers. Yeah,

Chip Rodgers  16:07  
I would think with that volume, that's got to be a big part of it is just getting the field to understand what the new capabilities are, what each partner, you know, with 1200 plus partners out there with agentic AI capabilities. How do you, how does, how do the, how does the field even digest it?

Jagjit Dhaliwal  16:30  
Yeah, so right now, it's very much focused on working backward from the customer. So this is something you probably have heard a lot from AWS. We are very passionate about working backward from the customer problems, and every customer is at a different maturity curve when it comes to the Gen AI adoption, and based on their maturity situations, what use cases are they looking for and then helping them prioritizing those use cases, and then for those specific use cases, what are the agentic solutions which are available and how? How do we help field, identify those and give those sales place and collaterals for them, so that they can map those things and make it easier for customers to narrow down those lists so it everything goes back to the business problems, what we are solving So, and I think you, you started the conversation with the reality check. I think that's exactly differentiate between the customers who which has already started getting the benefit from it, versus customers which probably maybe already doing lot of experimentation. It all boils down to that how they are executing Gen AI as if strategic priority within their organization. If it is a very bottom up approach with haphazard lot of POCs and trials, you will see that customers may not have yielded a value, but if it is a strategic priority, working backward from a business outcomes which they want to drive, and with the exact sponsorship from the CXOs, that's where you will see the prioritization comes in. And this is where there you'll see that the enterprise readiness and the organizational readiness is there to go beyond just an LLM conversation and look up, look up Gen AI as the end to end solution.

Chip Rodgers  18:27  
How much? How much are the solutions, the different solutions you talked about, use cases, which you know, of course, that makes sense. That's a way to sort of organize and think, think about, think about the, you know, how to, how to put them in the right mind space. And I assume there are also many that are very industry specific as well. How much of that are you seeing? Where there really are, you know, instead of just sort of horizontal, LM, AI capabilities. So they're really getting into very specific industry use cases and building both technology, but then also the reputation within those industries.

Jagjit Dhaliwal  19:13  
No absolutely see, as I said, like customers are on different maturity pathways. What we have seen is customers who are embarking the journey, they tend to go with lot of proven horizontal use cases, some of the low hanging use cases, which I have seen a lot in the start of the journey, is customer experience or employee experience related use cases, help desk or customer desk, Help Desk kind of use cases, because that's where Gen AI can help a lot and make it personalized experience rather than a robotic experience. Then the software engineering, coding assistance is a big avenue where generative AI is already making migration of the applications be. Need developing new applications using the NLP based approach for the coding, that's the second big area, then sales and marketing. That's another horizontal area where we are seeing a significant success with a lot of marketing, content generation and collaterals campaign. So these are the low hanging fruits, which we see that from a horizontal perspective, where customers are already getting a lot of value. But as you rightly said, we have started seeing a trend around going vertical use cases now and usual suspect industries, financial services, manufacturing, healthcare, they're definitely a way in we're seeing use cases like fraud detection or risk modeling, customer onboarding in financial services, loan approval processes. So kind of use cases are there? Healthcare? Lot of patient related use cases, patient reauthorization, coming up with a very personalized experience for the patient interaction, then manufacturing, predictive maintenance, the supply chain optimization, all these were traditional AI use cases. Now we can actually take it to the next level by making it more personalizable and add the generation part of it, and summarization, part of it. So those aspects coming in, I think there are couple of industries which are definitely very fascinating, especially with Gen AI space. I mean, if you look at life science, regulatory industries, it's hard to implement AI traditionally in the past, especially when we are talking about the drug discovery or clinical trials, right with competitive AI companies are able to do the synthetic data generation. I think those are very complex use cases, and we have already starting seeing success in those areas. And media and entertainment goes back to my Cognizant TCS life. That's where, like, 10 plus year I spend in media entertainment, and it's so fascinating to see that how quickly that industry matured in terms of the creative content generation. I think there's no limit to it, the kind of imagination which we're gonna see in our next generation of content, which is coming in, and the animations and the effects which we are adding into our content, I think that's mind blowing. That's, that's another big industry I'm seeing that gonna get disrupt which probably wasn't there like 10 years back.

Chip Rodgers  22:33  
Yeah, yeah, it is, it is, it's, it's crazy. Yeah, if mind blowing is the right word, I think

Jagjit Dhaliwal  22:43  
it is fun. I think it's I mean, we can't even imagine all the impact which we're going to see here, downstream down the line. So,

Chip Rodgers  22:51  
yeah, so what are you seeing? You know, you're obviously working with partners every day. What are you seeing in terms of sort of end customer adoption, are we get? Are we getting, but you sort of touched on it earlier, but are we getting, really, getting past sort of the POC level kinds of things, and getting into, you know, really large scale deployments and strategic projects?

Jagjit Dhaliwal  23:21  
It's a very interesting question. Chip. That's why I'm a little bit smiling. See, there are two school of thoughts, you will, you will hear a lot of reports and the market trends, which talks about, there is a significant level of optimism around Gen AI, acceleration and adoption, you know, innovation, what we are seeing across that's kind of a testimony of it. Then you also will see the reports like and if you have read the MIT report, which came out a couple of weeks back, it was making headlines across the board, 95% of Gen AI pilots are failing. You will see these two spectrums, and you will see there is a truth in both and the way I'm seeing it is, I mean, if you step back and look at last two to three year journey, I mean, when Jenny I came in, this is the free tool, which we gave access to the employees and our customers and to the normal human beings. So there was a lot of experimentation started across organizations too, and be it called ad hoc POCs or pilots and so on. But if you look across any other emerging technology, it's not just Gen AI. If you have these kind of disparate bottom of efforts, it's good as excitement, but it's not going to yield the business outcomes which you want, right? And what we have seen AWS being one of the largest in our enterprise space, we have seen that how enterprise customers have really pivoted from this experimentation mode to the really production deployment mode since last. Tier,

Chip Rodgers  25:00  
yeah, that's great. And

Jagjit Dhaliwal  25:02  
we have started seeing that shift, and there is a set of customers which have started getting a significant value, and there are set of customers which are not and there were consistent themes which I was observing in all my conversations with customers and partners. And I think it boils down to three key big areas. So first one is enterprise readiness itself, the data and the integration and the enterprise architecture itself, because everyone has access to llms, that's not the differentiation, but the differentiation is, what's your enterprise readiness, and what's your data readiness? How do you make sure that you are leveraging your data to convert that generic AI to the generative AI, which is solving a business problem for your organization, which is specific to you, and every organization, I think, roughly, has like, 1000 plus enterprise applications with data sources across the board. So the traditional problems are still there. So Genia is just a one shiny object in that whole enterprise architecture. So before you even solve the AI problem, unless you solve your data accessibility problem, data quality problems, you're not going to get the value out of it. So that's one area of focus. And second area of focus is around the whole governance, security and compliance, especially in case of regulatory industries. Customers are asking question enterprise customers, they are specifically asking questions about how do I'm going to meet these stringent requirements of my regulatory industry, or the security requirements? How do I am going to get the model transparency or any concerns about the bias or explainability? So all these concerns are there, especially for enterprise use cases. So unless we address those, nobody's going to put solutions into the production. So you have to help with those conversations with the business lob leaders and provide them that transparency, that what risk appetite they are willing to take, and what it means and what it is not going to do. That's the second big bucket. And I think the third big bucket is the organizational readiness, and that goes back to any technology adoption. And you mentioned about the deputy CIO role, and that's something which I experienced firsthand in public sector change management, working backward from the business priority of the organization. And how do you not just swift away from the new emerging technology, and how do you first start from the business problem and a business outcome and then work backward from there, and then help with the bringing a clarity on the ROI? So all these are organizational skill set which are needed, and it require a different skill set if you are implementing the Gen AI technologies, whether you build that skill set in house, or whether you leverage the SI Partners to do that, outsource that, but irrespective, you need to build that muscle. So these three components, the enterprise customers, which has done that from a top down perspective, we are seeing that they are embracing these technologies and already started seeing the ROI coming out of it. But if you have a bottom up approach with disparate pilots and POCs are going in, and if you factor in that count, yes, I agree the percentage of 95 probably might be true, but if you count the true pilots going in as aligned with the strategic priorities, the percentage of POC to production has definitely made better than that. Yeah,

Chip Rodgers  28:49  
that's really interesting. That's a great way to think about it on those three dimensions. I think that's that's a great sort of takeaway from what you're seeing in the market, both with partners and also with within, within customers. And I love that you brought in your experience as a CIO because

Jagjit Dhaliwal  29:10  
it was, it was, I mean, I would say that's one of the best experience which I got. I know you touched upon the different organizations. I did not plan it that way. But in hindsight, I think the one thing which I really feel proud of is that I got a chance to work in different geographies, different industries and different phases of the software life cycle, be it a hunting, be it a farming, sales, be it a delivery, be it now partnership, or be it a strategic technology planning. So as a deputy, cio i was leading a technology strategy for the whole LA County for data and automation AI space. And just to give you a size of the organization, LA County is the largest county in us and. The the county budget is roughly about 35 billion. So our IT projects was a size of, like, easily Two $300 million projects, and overall budget, like in the range of 101 to 2 billion. So and 40 plus different organizations, which are very different from each other. Like the police, Sheriff services is very different from, let's say a child services or or let's say the public works. Every department is a different organization in itself, right? Imagine if you have to come up with a top down AI strategy for those organizations. So that is where the change management and how do you lay out the whole enterprise architecture approach? And working backward from a business is extremely important. You cannot push the technology and assume that it's going to happen. Yeah, and all other I mean, after doing that for that county, got a chance to lead the CIO practice for UiPath. We're got a chance to implement same things for smaller scale customers, but you can see the patterns are still the same. As long as you are fundamentally approaching it right way, from the top down and bottom up evangelism, you'll be all good.

Chip Rodgers  31:15  
This has been fascinating. JD, maybe one, one other question, I'm curious about how you and your team, what is success for you and your team? How do you, how do you measure it? What's the what's your criteria? What are your goals, to the extent that you you're able to share them? I love to, love to hear that.

Jagjit Dhaliwal  31:39  
Yeah. So see, end of the day, we are partner organization. We our success is dependent on a partner success because we help our partner, co sell with us. So it's a joint success. It's not that aw is going to succeed and partner is not, and vice versa. So as I'm wearing two hats, if I look at the Jenny ISPs, where I have a direct responsibility of partner management, the joint courser revenue is one of the goal how, how are we helping our partner increase the GTM and drive that business revenue growth year over year? That's straightforward metrics, and since marketplace is one of the preferable channel for AWS, as well as for our partners. So we do look at how much of revenue is being driven from marketplace, and how are we making it. But then there are a lot of KPIs also, which we track. Because as a partner, we look at the whole life cycle. It's a build market answer. So it is about how we are co innovating with our partners to develop the solutions. And then how are we GTM in those solutions? How are we helping with the demand generation of market through the market campaigns? And then co sell is where we are going, very targeted, and helping them close those opportunities jointly. So there are a lot of KPIs comes into picture. Number of joint solutions we have built. What type of AWS services are being leveraged? And since I am from genai ecosystem, I'm always have a bias for how many AI services have been used, in addition to all other AWS services. Sure, then marketing campaigns, there are a lot of AWS led campaigns, and there are partner led campaigns. And how are we helping together and generate the market qualified leads around it? And then when it comes to the specific opportunities, how do we help jointly on account mapping, identify the pursuits, because AWS is global company. We have, like, millions of customers. But how do we ensure that we narrow down to the pursuits which partners care about? How do we identify that by region, by industry or by type of customers, and help them identify the ICPs and then go after those specific opportunities and help them co sell there. So those are the KPIs and metrics around that. And then when we look at the overall ecosystem from genai lens, then it is about the KPIs around that. How is the genie adoption happening across ISVs? A lot of soft boots can work around that. If you think about examples of a business application ISVs, they may have a revenue goal, but is Gen AI is able to accelerate that revenue goal. I mean, with and without Gen AI, how is that matrix is coming along? And for that, like we have a different motions. I mean, one of the one of the very specific motion which our partners love is our strategy collaboration agreements. We have a Gen AI version of strategy collaboration agreement as a part of which we invest in these three pillars, build market and sell, yeah, and then accelerate that journey. A and we track that, how how we are investing, and what type of ROI is partners are getting from it, and what type of different motions they may need as a part of those SCS. So that's a feedback loop which we constantly keep evaluating. So those are some of the metrics and KPIs, along with the broader goals, which my team is being measured on. But one thing I probably missed, and I do want to call out, since this is a new space, customer stories is extremely important the success, because goes back to there is still a cohort which is like kind of laggers. They haven't onboarded the journey, and they look for other customers to really see that how their journey happened. So through our marketing campaigns, through our webinars, through our case studies and publications, we do bring in those success stories, highlighting what challenges they faced and how they overcome and what type of best practices and learnings came out of it, so that we can make it accessible to all other next set of customers which are embarking that journey. That's another area. I would definitely call it out.

Chip Rodgers  36:12  
So that's fantastic. I love that. You know, I think that's it's something that I think everyone I loved hearing here. Thank you for sharing that. JD, and I loved hearing that each of those, each of those KPIs, those goals, something is common across all very aligned, right? It's, it's aligning AWS with the partner for customer success. That's terrific. You know, it's, it's really, it's clear that that AWS is very partner centric, and your arm in arm with partners going to market together.

Jagjit Dhaliwal  36:52  
Yep, it's, as I said, it's never a dull movement, especially in this category, every day is a learning experience for me, and a day I don't spend looking at a news you never know what new model comes up and change the whole leaderboard.

Chip Rodgers  37:09  
JD, thank you so much for taking the time today. Really appreciate you sharing your insights and all the terrific work that AWS is doing around Gen AI, this met and exceeded my expectations for our conversation today, and really appreciate you sharing.

Jagjit Dhaliwal  37:26  
No I appreciate honor is mine. I really enjoyed this conversation, and always love chatting about how our partners are making a big impact, along with AWS. Always, always love these conversations. Thank you. Thanks for giving opportunity

Chip Rodgers  37:43  
Absolutely. Thank you. JD and to everyone else, thank you all for joining another episode of Inside partnering, and we will see you next time. Thanks everybody. Bye.

‍