How Can AI Improve End to End Visibility for Supply Chain Leaders?

Watch the recorded webinar featuring Zubair Magrey, CEO & Co-Founder of Ergodic AI, as he explores how artificial intelligence is transforming end-to-end supply chain visibility for enterprise organizations.

Designed for supply chain leaders, operations executives, and logistics professionals, this session highlights how Ergodic’s AI-powered supply chain platform delivers full lifecycle visibility, uncovers hidden risks, and enables faster, more confident decision-making.

In this webinar, you’ll learn how AI-driven supply chain management helps organizations improve transparency, optimize performance across planning, execution, and monitoring, and build resilient, scalable supply chains with measurable ROI.

Transcript:

00:01 Welcome, everyone, and thank you for joining us for this webinar today. Our session is going to be focused on one of the core challenges facing every supply chain visibility. 00:12 As the past few years have taught us rather growingly, supply chains aren't failing because teams don't work hard enough. They're failing because decisions are made with incomplete delayed or fragmented data. 00:24 And when visibility breaks down, predictability and resilience break down along with it. Today, we'll be exploring why traditional systems can't keep up with today's complexity and how AI is changing the way organizations detect risk, act faster, and protect performance. 00:42 To guide us through this, I'm excited to welcome Zubair, CEO, and co-founder of Ergodic AI, who works closely with supply chain leaders tackling these exact challenges every day. 00:54 With that, let's get started. Nice to have you, Zubair. Thank you, Michaela. Hi, everyone. Thank you for joining us today. 01:01 Really appreciate your time. We have a very jump-packed agenda which I hope to proceed through our good pace. Today we're going to be talking, as Michaela said, around the end-to-end visibility problem in supply chains. 01:15 We'll take a little bit of a tool. We'll try and understand why we have the problem, why existing tools and capabilities don't quite tackle and provide supply chain leaders and operators with a tool and that they need, how we are a Godic are looking at this and how we try to solve it and what it actually 01:32 means when you are able to solve into invisibility, what new capabilities are you gifted are you given and then there'll be time at the end for a Q&A as well. 01:44 So without further ado let's get going and start talking about the visibility problem in supply chains. So let's start with some very recent statistics from a survey undertaken by Reuters at the end of 2025. 01:58 Essentially questioning several hundred supply chain leaders across industries on the challenges that they're facing both externally and the key initiatives that they're looking to implement internally in order to meet these challenges head on. 02:16 So figure one gives us a sense of the major external threats ranging from geopolitical tensions, demand supply and balances, natural disasters and of course they're all from the pandemic. 02:29 So there's a whole bunch of them right which are really outside of any supply chain leaders control. And then when we look at the initiatives, the kind of priorities if you like of supply chain leaders, a number of them come out which resonate with a lot of the conversations I have with leaders in supply 02:48 chain. So cost reduction and the desire to do more with less is a fundamental priority across every supply chain leader. 02:58 Then it's looking at how we tackle these disruptions, so resilience to disruptions. And then in third place, or a very close third, is the ability to enhance supply chain transparency and visibility. 03:13 And it's are pointed view that actually in order to tackle the cost reduction challenge, in order to improve resilience against these external challenges, and actually in order to adopt a number of the other initiatives at the bottom, whether that's improving relationships, driving increased delivery 03:29 speed and efficiency and so on. Understanding into when supply chain visibility is absolutely central to it. But let's first of all define what we mean by end-to-end visibility. 03:43 So I'm going to do whatever one says not to do which is read out the definition here. But what we mean here is end-to-end visibility is knowing in real time and on a forward-looking basis how every supply chain evental decision propagates across the network and ultimately how that affects the metrics 04:02 that matter. So whether that's customer service, cost and risk, and the final point and this is really crucial is being able to act on it. 04:12 So not just notifications, not just another dashboard. We want to really activate this end-to-end visibility. So we navigate these disruptions, navigate this cost reduction initiative, navigate whatever key challenge or key objective we have our supply chain. 04:29 We want to be able to use this end-to-end capability to help us navigate that path forward. But why is this hard? 04:38 I think the diagram on the right hand side provides a lot of the corporate reasoning once the way it's really hard. 04:45 Across every large complex supply chain, there's a number of different systems of record which help to manage different parts of it. 04:57 So product side and you're looking at a PLM system, you might be in manufacturing and are using your asset management system or your ERP in order to run your production schedules. 05:09 You typically have some kind of BI reporting tool which sits on top of these different systems to allow you to report on maybe production adherence or OTIF or another metric that's important to you. 05:21 The reality is though that each of these systems they tend to be hugely disconnected and basically siloed. So in order for us to understand or answer questions that perhaps straddle the ownership of these systems, we really have to then end up pulling extracts or pulling reports or merging different 05:41 reports together, often in a really manual and time-consuming way. So that's going to challenge number one, which is this complete disconnect of systems. 05:54 Challenge 2 is actually around the real time nature of what we need to be able to respond to. So where there's, for example, a new tariff that's come online or where perhaps a particular workstation or plant has gone offline, these can be real time signals that we can react to and change course and take 06:20 an alternative action or take a remedial action. but oftentimes real-time data is completely outside of the loop of anything to do with him to invisibility that it leaves a lot of a lot of value on the table. 06:33 We also have issues around limited forecasting or really basic forecasting, so the vast majority of supply chain leaders I speak to, they struggle with generating an accurate forecast and some of the reason behind that is this lack of connectivity. 06:51 The data is always siloed, maybe it's too difficult to get all the data into one place together and therefore to be able to run accurate forecasts which take in all the different viewpoints whether you're in ops, whether you're in the ERP, whether you're in your warehouse. 07:09 Being able to bring all these different elements together in order to understand the most accurate forecast possible is also a kind of big challenge. 07:18 So pairing all these together, that leaves a lot of stressors on the table for supply chain leaders because essentially instead of being able to interrogate a single source of truth around end-to-end visibility, we're having to go into these disparate siloed systems in order to get or try and piece together 07:39 and answer, and that in and of itself is highly manual. So you have teams of data folk teams of business analysts teams of supply chain leaders who are spending all their time just trying to piece together. 07:52 All of these different data points to answer questions from above. And that's just the questions internally right so if we also are facing external challenges which are really putting stress on to our supply chain network. 08:05 it's almost impossible to react to that with this level of disconnectivity and that's kind of why visibility is such a hard problem to to crack. 08:18 Taking this step further, again when we talk to supply chain leaders, operators across industry, we hear the consistent set of needs that they want to achieve, right? 08:30 What is it that we want to get out of being able to have this end-to-end view or when visibility across our supply chain. 08:38 And these five come out repeatedly. So visibility is obviously one, it's not a nice to have, it's a must have. 08:46 But on top of that, what we want to be able to do is to identify risks both ahead of time and being able to respond to those risks in a timely fashion before those risks cascade into these four-blown crises. 09:02 Another way to think about that is how we describe different types of insights. So we want to be able to predict something's going to happen. 09:11 This is more on the predictive side, so is an upstream supplier issue going to cause an issue to a downstream customer who happens to be a VIP that I care about? 09:24 Is that a true action? So that level of prediction is really important. But we also want to move beyond that into more prescriptive insights. 09:35 So this is really the action part of that end-to-end visibility definition. Because if we're able to understand in a prescriptive fashion what we should do, we as humans, we as supply chain leaders can then act with the augmentation of really clever AI and really clever intelligence instead of having 09:55 to spend our time and piecing together these different disparate systems in order to work out what to do. The fourth point here is around agility and being able to respond to the myriad of changes that are both external and sometimes even internal as well. 10:15 There are many internal challenges as I'm sure you're all familiar with and being able to piece together the right responses, both internally, externally But also with our multi-tier supply chains is also really important. 10:28 So this is not just an internal, how does our internal supply chain work? It's also how do how do my suppliers and their suppliers and how do the issues that they face impact my ability as a supply chain? 10:42 It is to deliver to my customers and to meet the metrics and the KPIs that I'm being held accountable for. 10:49 And the final point which has come up more recently is around end to end orchestration. So there's a lot of talk around automating workflows and automating the end to end and orchestrating actions across the end to end. 11:03 There's not a chance in hell that you'll be able to orchestrate your end to end supply chains if you have the silos of data across your different systems through which you run your supply chain as just impossible. 11:15 And that's the reality. Data is fragmented. Issues when they arise often have already become a crisis and we're having to go back and retrospectively root cause the issue in some kind of war room with another 20 or 30 people from across the business. 11:34 Insights tend to be really descriptive, which means backward looking. So every operator and every leader that I speak to has dozens, if not more reports and dashboards through which they and their teams run their business, it's not the same as having the right recommended actions, these prescriptive 11:52 actions that you can take. And finally, the amount of time and effort spent on manual data crunching, non-crunching type work as opposed to overseeing end-to-end orchestration of key actions that can and should be automated with human oversight where possible. 12:13 So ultimately, to conclude this point, traditional systems were not built for visibility. They were built to run a specific vertical in the operation. 12:23 Therefore, this fragmented view, rather than the complete picture, without the complete picture, we're left fighting fires retrospectively, which is not an ideal place to be. 12:35 So, a little bit now about agonics. So, when we started working in supply chain several years ago, one of the first elements that we felt was table stakes was to try and solve this end-to-end visibility problem because it enables so much of the more intelligent automation orchestration work that we want 12:58 to help our customers towards. So what it means in anger is the agonic platform will connect into all of those disparate systems and piece together the end-to-end chain of entities like suppliers like works and there's like products, DCs, even customers. 13:18 And it fills the connectivity with the events that drive value. So, what are the events that go or help us take a raw material or component coming into our supply chain and ultimately converted into a higher value product or output that comes out of it? 13:37 So, every event ranging from a new supply coming in, a new forecast, a new production plan, a production run, the movements of goods, routing every single movement and every single event is meticulously trapped throughout. 13:55 So that's the kind of high level approach. And then what you see on screen, which is what I'm going to demo for this audience in a second is how we essentially think about making that visible to our customers. 14:10 So what we wanted to do was to have essentially a single pane of glass that sits across all of those entities and all the entities in your supply chain are both internal and external and allow you to essentially almost swipe left and right. 14:26 So if you want to understand what raw materials went into a production rank and if you're in a production centre you should be able to swipe left and go and see the raw materials. 14:43 If you swipe left again and see the suppliers. But Ditto, if you want to see how a particular raw material ultimately ends up as a finished good and in the hands of a customer, you can swipe right and move through the supply chain. 14:55 So we can see every input, every activity and every output. And that's how we build the basis of this end-to-end visibility layer. 15:06 But as a reminder, end-to-end visibility is not just connecting the dots as being able to activate it. and that's what I'm going to show you in the demo that I'm going to take you through in a second. 15:17 So let's go, let's get into that. So what you can see on screen now is the agoric platform and specifically we're in Atlas which is our end-to-end visibility capability. 15:31 All the data that you're seeing on screen is synthetic or generated so don't read too much into the fact that we're looking at double eye needles here, but just assume that this is a product page, so it's a third. 15:45 And every page has a broadly similar structure where there's a central panel which contains all the related information to do with that specific entity. 15:55 So because it's a product, we can see that, for example, this particular product has very poor out of 26.3% and we can see some of the other associated top level metrics around it. 16:09 And as I said before, end-to-end visibility is not just connecting the data, albeit that's a big challenge. It's also being able to activate that data with intelligence that ultimately can be used to drive better performance. 16:24 And one example of that is root cause analysis. So we also alternate root cause analysis with our with our engine. 16:33 And we make that available at every entity page. So whether you're looking at a particular product as we are here, whether you're looking at a supplier, whether you're looking at downstream at a customer, we can tell you the reasons why we see failures. 16:49 And in this example, we're really talking about OTIF failures, so the ability to deliver on time and fall to the customers. 16:56 But any metric that you care about, we can deconstruct it into these, what we call these root cause chains and that's just one example of the kind of intelligence that you can start then applying and showing directly into Atlas and essentially make that available to your product managers, your logistics 17:15 managers, essentially your operating leaders who want to then use that data to drive improvements and naturally if you know that for this particular product we often are facing missing inventory and capacity constraints. 17:31 And if this is particularly valuable to you, then you can spend the time, resource, and effort to hopefully try and fix this root cause so that we don't see it repeated in future, thereby improving the efficiency of your supply chain. 17:43 We also look at events when we think about visibility. And this was the connectivity, if you like, between entities. And we track every single event down to its microscopic Scopic level of detail. 17:56 So every time we've seen missing inventory here, we've got full tracking, including the number of times, the quantity, the dates, and so on, which again allow you to then, instead of spending time, for example, trying to understand why things have failed, you can come in here and you know everything 18:13 that's gone on with this particular entity. As I said, Atlas is a single pane of glass, so we can go backward and forward. 18:21 So backward from product, But we can look at the different semi-finished goods that went into this. On this example, there's nickel and nickel-plating solution. 18:31 We can see the production plants that were involved with producing this specific product. And we can also see the work centers involved in generating. 18:40 So these are the work stations, the machines and so on. So maybe let's click on to a production plant for now. 18:47 This is plant 004. And we can see that actually there's some some interesting stats immediately. So we can see that there's a different root cause chain here. 18:56 So we can see there's different issues that perhaps we're seeing at this part level that we want seeing at the product level. 19:03 And again, different personas, different leaders in your supply chain can start them activating this and start making making the appropriate improvements. 19:13 any metric that you care about, so that is important. We also show, so, for example, at a workstation level or a plant level, we're interested in the ability for us to actually produce against plan. 19:26 And therefore there's a lot of metrics here around how does this plant be able to actually exceed or at least meet the plan. 19:33 And we can also see all the different materials that are involved at this particular plant. Now let's take a couple of examples downstream, just to show you the power of this. 19:43 So let's go back into the cruelty needles pack and it's the same, I think this might be a different product, but nonetheless we can kind of see what's going on as the same view. 19:52 And then when we go downstream we can then see, once you have developed or manufactured that product, in this example it then moves into a distribution centre for onward logistics. 20:04 When we move into the DC, we can then see all the materials and all the all the related statistics and insights for this particular delivery centre, including where we have issues in the past, material history, productivity and so on. 20:22 And then downstream from the delivery centre, we can actually see which customers this delivery centre delivers to, and I'll just click into one of them for now, it doesn't really matter a customer zero nine and then we can see different stats, customer order analysis where they see failures and so on 20:41 . And that's in a quick five minute tour. What Atlas is, it's to repeat, it's the erotic way of solving end-to-end visibility. 20:52 So, connect your entities with the events that drive them and enable intelligence on top of that, whether it's a root cause and access, whether it's trying to find and surface the actions, it can all be done with the single pane of glass. 21:12 Now I want to move on to what this enables and I've touched upon some of it but there's there's much more. 21:18 So once you have this end-to-end connectivity and into invisibility in your supply chain, there's a number of different ways you can take this. 21:28 So I've talked a little bit about the data foundations or I'll uncover that off here, but with respect to intelligence. 21:35 So we can start identifying patterns. We can see, for example, when a particular supplier, when they have issues, which downstream customers that fits. 21:47 So if you have a really strong procurement functioning, you need help in keeping your suppliers or getting on top of your suppliers, it's the perfect tool because you can look at a particular supplier and Steve Bage and see when and where and to what extent and materiality they have impacted customers 22:04 and see the quantified impact to an end customer. That's really powerful. Ditto if you want to understand perhaps how different plants are performing and the impact of those plans on a customer service or a particular risk type is really straightforward to be able to answer those kind of questions. 22:26 It also provides the standardization as we saw with the way in which the pages on the navigation instruction were able to then whether you're a logistics manager or a plant manager or a customer manager, you can essentially see the same view and be held honest with how you govern and how you manage your 22:45 particular domain, particularly if you're getting impacted by another part of the business or if some of the issues that you're even within your particular domain are impacting others, it's all there, it's all connected. 22:59 There's no hiding place, but then by the same token, there's no blame being because what we're trying to do is improve everyone. 23:05 So the rising tide that raises old boats With this end-to-end visibility, we're also able to then improve the inputs and outputs across every web of our supply chain, which is really powerful. 23:16 The final two bits here, so operational analysis and strategic optimization. This is really where you then start to see the benefits of having end-to-end visibility. 23:28 Being able to, for example, I gave you a couple of examples with the demo of how the root cause analysis capability can be activated in different entities in different parts of your supply chain to fix those recurring bottlenecks or issues. 23:44 That's a core capability that you would not be able to have or see without having this end to end visibility. 23:52 But it also gives us the ability to run really clever scenarios. So we're able to understand cascade both forward and backward the likely impact of maybe an external disruption. 24:07 So for example, if there's going to be a weather issue or a some kind of trade union strike at a particular port, we can quickly simulate the impact of that because we have the suppliers connected into every single part of our network as we can see without visibility. 24:26 So we're able to then run some of our reason, and ideally build resilience into our supply chain as a result of understanding how cascades actually occur. 24:37 So it's an incredibly powerful capability that is transformative for those who are willing to put the effort to adopt this capability. 24:48 And that's kind of where we are today and that's the quick end to end of end to end visibility. We now have a few minutes left for any questions, so Michaela, over to you to see what questions we've been sent through. 25:04 Yeah, of course. So we had a few sent through during the slides. The first one that seems to have come up was, how is this different from a control tower? 25:16 Yeah, very good question. So we see control towers kind of everywhere. And the way I think, the way we think about control towers are a collection of nice screens dashboards, which are still really siloed. 25:32 So they are still fundamentally affected by that great diagram that we had up front in this webinar, which was all the different siloed systems. 25:44 So the control tower really takes, might take a report from your warehouse management system, another graph from your ERP, another graph from your POM, and puts it all together onto one big set of one big screen. 25:57 But essentially what you've got is a reports or a dashboard here and another dashboard there, it's all in one environment. 26:04 So number one, they're always really backward looking. Number two, the connectivity or the visibility is somewhat false because it's not really connected data is a series of reports that have just been stitched together. 26:19 So when we talk about back to the definition of end-to-end visibility, the desire to be able to act on the insights as a result of having this connectivity, control towers really struggle with that. 26:33 They're really descriptive, rather than being particularly strong predictively and certainly not prescriptively. And that's the difference, it's, do you want to have visibility, disconnected visibility looking backwards, or do you want to have four end-to-end connectivity and visibility and the ability 26:52 to simulate forwards and understand backwards, and that's what we are trying to do at a gothic. That's great. I actually did know that myself either, so that was a great question. 27:05 Another one that I think a lot of people may have is what can you do if you only have real-time data accessible from parts of your supply chain? 27:14 So they said for us we have some manual forms and documents that are not always able to be processed all in real time. 27:22 Yeah, it's not something to worry too much about. The way the way to think about it is this not every every single, not every single entity or every single event needs to be available in real time in your end, this end to invisibility layer. 27:43 And actually it might actually create too much for teams to have everything real time. So if there's only a certain number of entities or events that are available real time, only take it through if it makes sense. 27:59 so we can be pragmatic about it. It's not, as I said, it's not always necessary, but if it is necessary, then put the right investment into the infrastructure, the connectivity, so that it's all fully integrated in our proper IT safe way. 28:19 But yeah, I wouldn't expect any organization to have every single data point available in real time. It's just really difficult to get to that. 28:33 Right. And the final question that just came through that's really interesting is how long on average do you see it taking your customers currently to surface issues without AI versus after? 28:48 Yeah, it's a great question. And if there is, we have some examples where where an issue has cascaded, for example, into a big outage or a big customer issue, where it's a complex issue, a multi-step, multi-tier, maybe a whole bunch of different reasons around it. 29:13 We've seen it take anywhere from weeks to even a couple of months to actually get under the covers and understand the why behind the issue. 29:22 All the way to some of our customers who can avail of this level of automation that we've built into the platform, where we're able to actually surface issues before they fully cascaded, literally live. 29:38 And it goes into maybe the previous question around real-time data. Naturally, the more real-time certain events are the more quickly we're able to identify, trap them and work out where and what we can do about them in order to remediate them. 29:55 But yeah we can go from weeks and months to identify issues all the way down to literally minutes and seconds. 30:02 It's that however on it takes an operator or a leader to click through one or two screens so it's quite an order of magnitude change that we can affect here. 30:16 Great, that was fantastic. That's all you have time for. You'd like to close us out. Yeah, thank you very much everyone for joining and taking the time today. 30:27 I hope you found it useful. I'm very happy to talk again if this is a capability of interest. Feel free to reach out either to myself, to Michaela, or have a look at the website and get in touch, but would love to hear from you. 30:41 Thank you again for your time.