Dr Ajay Dholakia: 0:03

Ultimately, this digital supply chain concept is becoming a differentiation lever for us, and for anybody who chooses to adopt it, you know and go after it from that perspective.

Tom Raftery: 0:19

Good morning, good afternoon or good evening wherever you are in the world. This is the digital supply chain podcast, the number one podcast focusing on the digitization of supply chain. And I'm your host, global Vice President at SAP. Tom Raftery. Hi, everyone. Welcome to the digital supply chain podcast. My name is Tom Raftery with SAP and with me on the show today I have my special guest, Ajay. Ajay Would you like to introduce yourself?

Dr Ajay Dholakia: 0:46

Yes, indeed. And hello, Tom. Good to be here. My name is Ajay Dholakia. I'm with Lenovo, data center group, CTO for our alliance with SAP and also Chief Technologist for our software and solutions development organization within our worldwide teams. So looking forward to the chat today, we have had a long standing partnership with SAP on various fronts. And part of my role, when I say Alliance, it means 360 degrees. So we are partners in developing joint solutions. We are buyers and our suppliers to each other. So that way, it's a multi faceted relationship going back a couple of decades plus. So looking forward to continuing, you know, leaning on that. And then let's see in our dialogue for some of the key topics Top of Mind that we have today. Okay, squared, C five.

Tom Raftery: 1:50

Now, can you talk a little bit about your own supply chain Ajay because I understand it's quite considerable.

Dr Ajay Dholakia: 1:55

Yes, indeed. And part of the Lenovo, I mentioned, I'm part of the data center group. And Lenovo as an overall group has other business units, intelligent devices group being the and one of the other ones main big one, which has in it the tablet, the PC, the workstations. And also we have the mobility, or the mobile devices group, which is Motorola, and other Lenovo branded phones and related infrastructure. And then there are, you know, business units focused on digital intelligence or on commercial IoT and things of that nature. So it's a large breadth of, I would say, hardware infrastructure all the way from something on your wrist, like a watch, to the phone, to tablet, to laptop, to desktop, to workstation to servers, in your data centers, it's a huge breadth of devices and hardware that we manufacture and supply to customers worldwide. I sit in the data center group, which is focused on the infrastructure going to end these days, not just data center, also to edge and remote branch offices, and so on. So that whole definition of where the infrastructure is sitting is evolving. But our supply chain is leveraging the company wide needs for parts and partnerships, right. So we know when we go to get supplies for a server, we are not just doing it for that server, we are doing looking at the entire scope of all what we shipped. And just to give you an idea of Fun Fact is, in a year, Lenovo ships about 100 million plus things. So that breaks down to three to four a second. So that kind of gives you an idea of the scale at which the supply chain needs to be operating, and to end across enterprises. And as well as you know, consumers.

Tom Raftery: 4:19

Okay. And there's obviously enormous challenges in that. And this is the digital supply chain podcast. So can you speak to some of the digital technologies that help you overcome some of those challenges? Or maybe identify some of the challenges first, and then talk about some of the digital solutions that help you overcome those challenges?

Dr Ajay Dholakia: 4:39

Correct? Yes, yes. So as you can imagine, the supply chain is basically our partners, suppliers on the one hand, and then Lenovo and then going forward through the distributors through the retailers to the end user customer. So it's a large network of participants. And in each stage of that supply chain, things are being ordered, and things are being delivered. And these are physical goods, components, memory, CPU, you name it, right? So what we need to be dealing with is, is the flow of parts necessary to put together the system or the device in place at the right time to be able to meet the demand that is coming. And while we can see that, yeah, that, if you have set up properly, this flows, the reality over the past, you know, decades has been, a lot of it is still human in the middle or human centric, manually handled steps where each recipient has to verify is did they get what they ordered? Did it come in on time? Is it properly functioning or damaged? Should I be able to move this along to the next stage in the supply chain network. And often this requires people calling each other people using spreadsheets to keep track of part numbers and serial numbers, even yellow sticky pads, to know that favorite person in the supplier partner, who they have built a relationship with who they can call and get things done. But then when this happens to be the week, they are on vacation, and so maybe I have to either spend time hunting for an alternative or you know what I actually would rather wait. And so you can imagine how delays come in how data dependency comes in, because there is pairwise data and reconciliation but network while there are issues and challenges. So those are some of the key challenges that of course get amplified when you are working, working to supply goods in 180. Markets globally, with the number of devices that I mentioned, how many facilities are there, how many suppliers, you know, we work with 2000, plus suppliers, and a huge bill of our supply chain budget in multiple 10s of billions of dollars going towards acquiring those goods and parts and so on. So that sets up the challenge. Yeah. And it's not that, you know, that's the only way it is operating over the year every so we have very dedicated supply chain engineering teams working to optimize every step. So there is been over the years and decades lots of local optimization, people have improved using it tools and so on, where I believe we are at today is bringing in some end to end technology platform platforms, I call them platforms, because they are enabling this dialogue or this flow of things that has to take place. So an example of course, is data reconciliation, an example is very finding or validating the exchange that is taking place. So this comes in with things like analytics, it comes in with things like IoT and edge centric devices, aiding in monitoring the flow of goods that is going on, certainly in monitoring manufacturing quality, right. So there so so now you get into well, potentially AI machine learning can come into play. And then there is the whole transparency and trust aspect, which brings in one of the other buzzwords of the day, which is blockchain or enterprise blockchain. I tried to separate that from the public blockchain that he gets excited about. But behind the scene in the enterprise world, there is a lot of attention being given and we are an active participant and leader in deploying those technologies. So so that's a breadth of things. You know, there is there is data and analytics, there is AI machine learning, there is IoT and edge related monitoring and quality assurance and things of that nature. There is blockchain for traceability, tracking auditability certainly there is also underlying or behind the scenes reliance on some sort of cloud technologies in in connecting these dots. So, in my mind, these technologies are again, enabling platforms. They don't solve the full problem, but they are indeed A key component of putting a solution together.

Tom Raftery: 10:03

Okay, and which one would you say gives you the best bang for your buck?

Dr Ajay Dholakia: 10:41

That's a really good question. And the reason is that it depends on where you are looking right? You know, if you go with our magnifying glass across the network, what is the problem that we are trying to solve? So here is an example. We do, for instance, from our manufacturing plants or from our suppliers, pallets of goods come into the warehouse, right. And sometimes they are already paid for and are Lenovo owned, sometimes they are still partner owned inventory supplier on inventory, until the time that we decide to bring it into our books and say, Okay, now we can open it in and make use of it. We have deployed video based technology to help track and automate. With the front end the video cameras distributed throughout the warehouse, and on the card that moves those pallets around. to automatically know there are two dimensional barcodes or dm codes put up on those. And so there is automation bringing, being brought in from that perspective. And then the back end, for example, in an SAP software is capturing all the data. So the connectivity is put together in that way. And now you have the ability for a person to monitor and just kind of keep an eye on where there might be exceptions. So things if they are flowing the way they are supposed to flow, the human involvement is taken out because the barcodes are checking, did the pallet go in the right aisle, the driver of the vehicle that brought it there, make a mistake and go to a wrong one, or did it get transferred between one supplier to another or between them and Lenovo, all this is happening automatically because bill of material is available can be attached to that barcode and then you can make that. So that's you know, an example of a technology that can really help automate and increase trust in the data that we are collecting and using for analytics. Another example would be I mentioned blockchain. So this is the notion of being able to trace and track it by necessary necessity, it involves multiple partners working with each other. And the whole, if I break it down, the key element is that it's a shared ledger, which means when transactions are recorded in the ledger, all participants have access to that record. And it's also immutable through its design. So once you record something, you are guaranteed that it won't be tampered with, or modified by somebody along the way. So with those guarantees, the transparency of the data, and the speed with which it is accessible. Both improve, right. And so now the speed of business goes from months and weeks, like we were saying earlier with human in the loop to potentially days and hours. And if we are able to achieve that, you can imagine how much of an impact it has on the return on the ability to contain costs, because exceptions are being spotted on the ability to have the traceability that says oh, I'm now able to actually shorten the time I hold or shorten the time I need to validate and therefore, my books are accounting books are freeing up resources faster, right? And maybe I can take that and do something more productive with the freedom, capital and so on. So, you know, those are some areas where the scale at which we operate, this is having a direct impact. Sure, sure. And maybe the third thing to mention is in general, machine learning is now pervasive technology that is being talked about in our personal life on you know, we deal with it every day, certainly in the e commerce and enterprise to enterprise a business to business interactions. And you can hear about all kinds of algorithms that are being captured based on just watching how things are moving, capturing the data and creating some learning and rolls out of that. The traditional machine learning way of doing things Because we have more transparent and trusted data becoming available, due to other digital technologies being deployed, the quality of learning or the or the confidence in the outcome from that learning or the predictions become that much more actionable, and that much more believable. So there is sort of also a synergistic effect here. That takes place if you start to bring multiple of these technologies to work together.

Tom Raftery: 16:08

Sure. Sure, sure. Yeah. One of the issues with machine learning, though, tends to be that it is only as good as the data is that you feed it initially. How do you overcome that potential issue?

Dr Ajay Dholakia: 16:22

Oh, that's absolutely one of the key areas where not just us. But you know, scientists, and engineers at large within academia and industry are doing a lot of work. So we have data scientists on our staff who work on implementing data cleaning and data collection techniques, and models and algorithms. And so it is an ongoing improvement that needs to be brought in. So we are very aware and working consciously, there are some framework level or model specific techniques that have come into light, which can be deployed. But ultimately, the biggest thing that any Deployer of machine learning can do is be conscious of this very fact, which you said, which is, it's as good as it's only as good as the data that you that you feed into it. So we have to be careful about how we train those models, where those data's are coming from, are we actually able to triangulate across multiple parts of the supply chain. In this case, since we are talking in that space, you know, we have supply chain for multiple lines of products, can there be learning across them, because we may be looking for defect in one type of device, but may not be aware of what it would be when it is actually being put in a different form factor. So if those can be looked at or jointly or combine, you know, so these are the kinds of things that our data scientists and data engineering teams work on to be able to achieve. You know, it's an ongoing improvement. It's not a one shot and you're done. type of facility.

Tom Raftery: 18:11

Sure, sure. Sure. And you mentioned you have over 2000 suppliers, I imagine the number of customers must be orders of magnitude even greater again, I mean, you you sell to distributors, you sell to retail outlets, and you sell direct as well to consumers, as far as I know. So that's that that's a very complicated web to try and manage. How do you how do you go about that?

Dr Ajay Dholakia: 18:40

Yes, indeed, is 1 billion plus global customers is kind of the top line number. And this is because we are in a consumer business as well, right? The phones and the watch and the tablets and PCs and so on the enterprise benefits from having that access the enterprise side of our business and supply chain. Right. So one of the things we have had to think about is customer centricity. So how, you know, supply chain, from one some perspective is a back end, you know, under the invisible, it's not seen, it's invisible under the firmament, you know, kind of a thing that should just happen. But part of what is emerging is that the more attention is paid, first of all in modernizing, but also in taking it all the way to the customer. So it's not just that you deliver on the doorstep and you walk away, if you bring in the customer in the loop, then they begin to see the benefit of the their own supply chain, enabling better results for them. So this is where a lot of focus from our side from our supply chain teams is is happening is interacting directly with the customers. It ends up being collecting feedback, collecting and monitoring their experiences, their interactions, to the point where some of these customers turn around and come to the supply chain function for their next business challenge directly as opposed to going through other potentially sales and services. So they are beginning to recognize that my problem has a solution in some enablement are some advancement that will happen from a supply chain perspective. And so keeping track of that monitoring that building those relationships, not just with your suppliers, but you know, the retailers, and eventually the end customers is where, and you can imagine once that is the focus, you need data, you need learning, you need validation, you need, you know, a feedback, type, traceability and actionability of what might happen next. So, again, those enabling platforms of machine learning and of data analytics, and, you know, IoT comes into play,

Tom Raftery: 21:15

right? Okay. Okay, there's got to be kind of issues, then around trust and transparency, do you want to maybe address that,

Dr Ajay Dholakia: 21:24

that is a key part worldwide, right. So, particularly when we talk about electronic high tech devices and, and server infrastructure that we you know, storage, networking, and so on, all the way from materials from which these things are made to the end user in how they are brought in, assembled, manufactured, shipped, and supplied and installed and used and then retired, there is a whole host of need for doing the track and trace and for need for doing some building of transparency, which then leads to increased trust. So this is where working with suppliers, we engage in initiatives, that may sound like catchphrases, but they are real trust your supplier, or, you know, building in network of participants that are able to collectively put together the necessary practices for enabling the traceability, do we become a leader in creating a consortium of sorts to in some cases, participate with other. So this is also leading to where you work with who might be competitors in the market, right, because there is common benefit to be had for each other and for the world at large right in being able to enable this. So a lot of that is being I would say these are the second and third level effects that have come into play, because the underlying technology is enabling the ability to have that type of visibility, that type of transparency, that type of access to data, which is decentralized in the sense that all participants can be looking at it rather than be trusting from just one source.

Tom Raftery: 23:28

Right. Right. Right. Okay, Ajay, we're comi g towards the end of the podcas. Now, is there anything I've n t asked you that you wish I h d any topic we've not addresse, that you think it's importa t for people to be aware o

Dr Ajay Dholakia: 23:42

I think one idea that we can put on the table here is that ultimately, this digital supply chain concept is becoming a differentiation lever for us. And for anybody who chooses to adopt it, you know, and go after it from that perspective. So you can actually win in the market, because of the investments that you were made in the digitization of your supply chain or bringing digital technologies. You know, the transformation is, again, one of those buzzwords, but as we have been talking today, you can see how attacking each little part of it gives you a competitive advantage collectively. And so we have this very deep seated notion that this is actually a differentiation metric. How globally are we situated in the world, for example, in the current year, plus, we have seen the impact of the ongoing pandemic on supply chain disruptions and so on. One of the things our teams are very proud of is that they have set up facilities throughout the world. So it you know, when something happens in one region, it doesn't spill over into the other and thereby you have the ability to keep the supply chain going.

Tom Raftery: 25:01

Right. Very good. Very good. Ajay if people want o know more about yourself, r about Lenovo, or about any f the topics we discussed on t e podcast today, where would y u have me direct the

Dr Ajay Dholakia: 25:15

So a good place would be lenovo.com. And then underneath data center group. I will also send you a few links for white papers and some blog posts related to the topics that we talked about that You can add it to your posting and people will be able to follow through that.

Tom Raftery: 25:37

Excellent. Great, great do that peasy. I'll include them in the show notes and people will have access to them from their super Ajay. That s great. Thanks a million f r coming on the podcast toda

Dr Ajay Dholakia: 25:46

Oh, absolutely. My pleasure. Thanks, Tom. It was indeed fun to talk with you about this very important topic.

Tom Raftery: 25:53

Okay, we've come t the end of the show. Thanks, veryone for listening. If you'd ike to know more about di ital supply chains, head on over to sa p.com slash digital supp y chain or, or simply drop me an email to Tom Raftery at sa p com. If you like the show, lease don't forget to subscri e to it and your podcast ap lication of choice to get new episodes as soon as they're publ shed. Also, please don't forget to rate and review the podcas. It really does help new people to find the show. Thanks. Catch ou al