Name: Mitratech ClusterSeven Roadmap
Uploaded: 2026-08-11T15:36:39.830Z
Duration: 34 min 59 s
Description: Mitratech ClusterSeven Roadmap

Transcript for "Mitratech ClusterSeven Roadmap": Welcome, everyone. Thank you so much for joining the main session with Alex. We are now gonna kick into our cluster seven road map update. My name is Charlotte, and I'll be working behind the scenes with you guys, to help moderate this session. Before we get started, just a couple of quick housekeeping items. For the main session, as we run through the slides, you guys will be muted. Differentiator for this one, as we get towards the end of the session, we have got a collaborative moment with Henry and Wayne, and you can raise your hand, come off of mute. We'll have a live discussion, in that portion of the presentation. As you're watching, any questions that come up, please pop those in the q and a tab in the right pane of your screen. We will answer as many questions as we can, during this live session. And as an FYI, there are some additional resources, in the documents tab if there's anything you would like to take away from this session. Final note, this is being recorded. We will have the session out to you by the end of this week, and the presentation starts to go with it. On that note, I will pass it over to Wayne and Henry. They're gonna walk you through all recent enhancements, innovations, and everything that's next to come with me. Thanks, Charlotte. Thank you. So I'm Wayne Belle. I'm GRC product manager, and with me today is Henry. So I'm just gonna quickly take you through what the agenda is gonna be for today. So, Henry's gonna provide you a bit of a product vision. I'm gonna gonna step through then some product updates and then a road map and then a presentation on well, demonstration on, Power Query. And then we do have a little bit of a collaborative session at the end just to talk about AI, not only its usage within Cluster 7, but also around governance and control. Over to you, Henry. Good afternoon or good morning, everybody, depending on where you are in the world. Thanks for joining us today, and I hope you enjoyed Alex's session earlier. I'd say, just, I guess, repeating Alex's vision here or our vision, is looking to ensure connectivity between, all of the different, solutions we have, allowing where possible AI to work across across those. Those of you most of you on the call here will be using ClusterSeven on premise. We do have a a cloud version which would make, a lot of that vision easier. But, we as I say that the the session at the end where the collaboration piece, we are very interested in understanding how we can embed AI into, into the cluster seven suite and how we can enable that connectivity with the other solutions. I guess if we just move on to the next slide Wayne, the majority of I think I'm not I don't intend to to read this out, but I think the majority of people who think of plus seven think of EECs, increasingly we are seeing that, our purview is, has grown outside of EUCs into models. So for those organizations that don't are not building their own, model systems, a lot of people are leveraging the capability in class seven to be able to manage their model programs alongside their EC programs within the cluster seven application. And then one thing again which we are looking to, with we're increasingly be asked about is, and Alex mentioned this in the in the first session, is governing AI. So, yes, we are interested in putting AI into our technology, but also increasingly, people are coming to us and saying, you know, what how can you help us with managing and monitoring organizations are now looking to see these potentially as EUCs. And again, I don't want to, preempt the discussion later, but it'd be interesting to know how many of you are being asked by your organizations about AI agents, and really, you know, how how can we help you, enable the that that management and provide the transparency the organization needs around those as we have been doing with your ECS and models. So that that's, I guess, teeing us up for further discussion later. I'm gonna hand you back to Wayne who's gonna take you through, the bit of the road map and a little bit of a demo and product update. Thanks, Henry. So, I'm gonna walk you through some key features that we released as part of the thirteen point two release of ESM and the 06/2002 release of IMS. So the first one I'm gonna talk about today is discovery pipelines. So discovery pipelines was is actually transformational. So the introduction of discovery pipelines means that EUCs and models can now automatically go from discovery straight into either IMS or ESM. That very much depends on your requirements. So pipelines can be tailored to meet your specific needs. Hopefully, this means that discovery no longer becomes just a manual task. It becomes more something that's scheduled and just literally runs in the background. The next one I was gonna talk about is the IMS and ESM time out. So over the next couple slides, you'll definitely get a sense that we are modernizing the security within cluster seven. So the time out feature is one of those vital features, provides a configurable time out setting that users will be timed out when inactive. So for an example, their session could last for an hour. Once timed out, the user will present it with a screen that enables them to return exactly to the record that they were on. Okay. I'm now gonna touch upon the road map. Okay. So this is the current iteration of the cluster seven road map. I will be going through each segment, in a little bit of detail, so apologies if this is a little bit laborious. But the the now are the items, effectively, that are available from you for you to benefit from today. The next of the items that are currently in flight and later items that we see on the horizon, but they may need refinement or more customer interaction just to understand what problem we're trying to solve. So whilst I'm on this slide, I'm just gonna cover off some of the the items that are available now. So the backup review and approval enhancement is specifically designed to reduce bottlenecks when it comes to those key member staff when they're away and unavailable. This means the process will continue to run. IMS notifications in sorry. Admin notifications in IMS will allow admin to send immediate and planned notifications to member's team about specific items, such as potential downtime or planned maintenance. We are continuing to expand our web client administration. You'll notice definitely on a few other items that we continue to chip away at this. This is our continued journey. So we will be moving specific vital pieces of functionality away from the console over to the web client, and you'll be finding out more about that shortly. I am gonna skip over to next because that there are many access slides, but I will talk to you a little bit about what we're thinking about in the later. So the quality of life improvements is really about honing down in Cluster 7 specifically to what's important to you, our customers. So understanding pain points and how we can alleviate that. Expanding the expanding discovery. Sorry. Our club is session is, at the end. So definitely, as Henry alluded to earlier, we wanna talk to you about how AI will benefit you not only within Cluster 7, but also how you manage AI from a governance perspective. But, again, as Henry alluded to, we'll speak into that later. Expanding this discovery so we can better understand dependencies between EUCs and models and other connected items. This is definitely something we want to explore with you so we can understand the true size of your EUC and model estate. And finally, multi record workflows. So the ability to have workflows and views associated and connected to the hierarchy of record types. Currently, we do have a limitation with workflows where they only run on a single record type. So definitely our big ticket item for the current release is entry ID. So entry ID is really a framework that enables cluster seven to connect to modern authentication methods. So we do have three potential options available. One is very much you stay in on active directory as you potentially do today. Option two is the one that we feel most customers will benefit from. It's still you having an integration with active directory, but the authentication piece is passed over to a SAML provider. So if your SAML provider enables you to use multifactor authentication, this would mean that would now become available within cluster seven as well. Our initial adoption phase is only for entry ID. However, I would welcome the opportunity to talk to you if you have other SAML providers that you want us to connect to. The final option is something specific for organizations who are SharePoint online. That potentially means that you've already done the entry ID migration. So as part of our next release, we will be providing some technical documentation that covers all the options available. I definitely know that customers have been actively talking to us about modernizing our authentication. So role management. So this is another critical piece of functionality that we've moved over from the management console to the web client. As I mentioned earlier, we are slowly removing our needs for access to the management console. But as you can see here, now application roles and users and groups will be available actually within the web client themselves. So ESM maximum session duration, so very much built on the back of time out. This is a similar ilk. So this effectively is another security feature built with smooth functionality to time out. When a user exceeds the maximum session limit, they are simply timed out from the solution and presented with another screen that allows them to log back in to the specific record that they're on. And the final one I'm gonna talk to you briefly about is around Power Query. So I am surely gonna provide a little bit of a walk through on this. But just to cover this off, we have now integrated Power Query as a data connection. So So this means the types of Power Query connections are available. They are Power Query formula, Power Query function, and Power Query parameter. I'm now just gonna take you over for a short demonstration. Perfect. Okay. Hopefully, everyone can see my screen. Yeah. So what perfect. Thank you. So what I wanted to do is just from here, just explain the way that Power Query works today. So I'm gonna choose one of my EUCs and then go to timeline. And what I'm gonna first of all do is just provide you with a little bit of an example of how it will work. So I'm not gonna choose the most complicated example to start with. I did wanna allude to the fact of what's in data connections. So when I click into data connections, what you'll notice now is we have additional options available. Obviously, this is to do with power query formula. What you will notice in conjunction with the formula itself, we will be creating an o l e d b, which basically means all the connection strings and other bits of information are available. If I hover over it, it will also provide me a summary of the formula or the actual formula that's being provided. To actually show you that a little bit more detail, if I go to external links, what you'll see at this respect now is exactly what you see today. We'll see all the external connections that are available. And if I hover over one of these, you'll actually see the command that we're passing over on each one of the connections as well as what's actually happening from a connection string perspective as well. That's a fairly simple one. What I wanted to do, though, is just take you through one that's probably a little bit more complicated, if I may. I'm not gonna choose that one. Perfect. Okay. I'll I'll just give you a brief tour of this one as well. So now when I click on external links, so if I go back to data connections, you'll see here the full array of the different options when we look at Power Query. So Power Query formula, Power Query parameter, and also the, sorry, formula and formula. There's not one for function on this this particular connection. If I go to external links now, what you will see is anything that's been added or removed or updated. So as you can see here, this data connection has actually been changed. You can see that with a green piece of information that's in the description. That will carry along if there was a change in the command for any window for any reason or a change in the connection string itself. In conjunction with that, we also have the capability now to use risk rules as well, from Power Query. So if I go back to the administration screen and then look at risk rules and then look at the Power Query connection so this is very much about looking for specific dependencies for certain sources of data. So something like a name server as an example. You can create a rule here to enable you to understand all the items that are best better connected to it. If you'd like a little bit more of an in-depth demonstration of Power Query, by all means, reach out, and I'll quite happily walk you through it. Okay. Okay. So what I wanted to do now is just touch upon the way that we, speak to customers about things that we're trying to gather in Insightful. So this slide is very much about how customer feedback directly shapes what we build and how we deliver. So we're using the same methodology we used a number of times already within the solution. So it starts with gather. We pull enhancements, ideas from multiple real world touch points on an ongoing input into Cluster 7. So structured user interviews, support cases, and customer surveys all gives us a broad continuous view of what matters most to you. And then we go through prioritize. So that's basically the voice of the customer, scoring and ranking inputs through direct customer engagement, effort estimates, and stack rating. So we're balancing customer impacts with delivery feasibility. So we focus our investment in what delivers the most value to you. And finally, we're into deliver. So through high quality predictable releases, this means automated and manual testing, clear customer communications, and training to support adoption. The goal here isn't just to ship features. It's to make sure that you've got confident, informed, and successful when using them. Together we ensure your feedback isn't just heard. It's consistently turned into meaningful, well delivered outcomes. If there's something specific I've mentioned or anything that I haven't, please do get in touch. We'd also like to run more customer working groups with you to understand the problems that you're facing so we can keep help shaping the road map as we move forward. So now we're gonna move into the collaborative session. So what I want us to do is just take the opportunity to talk to you about AI. At this point, I would welcome any active engagement to talk to us about this this topic itself. We do have two phases of this. So I'm gonna start on the first one, which is about how we actually actively embed AI into cluster seven. So we already have solutions, as Alex alluded to in his keynote, that already have AI embedded within them. So the difference between them and our typical clusters typical cluster cluster seven customer is that cluster seven is actually on premise. So things like document summarization is a feature we already have available, but it makes it a bit more problematic when we look to ship that to you. How do we embed that to an LESU user? So how do we connect you effectively, an on premise solution, to our cloud? The document summarization is actually a stateless service, which basically means that we do not retain any copies of any documents, whilst the service is running. However, we appreciate the sensitivity of the documents and spreadsheets you have available. So one consideration could be we ask about you bringing your own I a AI to the table. So is that something that your your organization's already adopted? Do you already have an AI provider? Have you chosen one? These are examples of candidates that we're looking at. So by no means is. this definitive list of all the options we have available, but we'd welcome your opportunity to just chip in and make any other suggestions on that. So so so, Wayne, if I could just jump in here. Given we want this to be interactive, if, basically, within the within this platform, if you raise your hand, if everybody can just raise their hand, then we can just put you on to stage so we can so people can be interactive with this. So, as Wayne said, I think there's there's two sides to this, discussion. The top half are, candidate items, AI features we think that, well, from discussions with yourselves, and with with others, these are candidate items we think that, you know, warrant some discussion. So the ability for every managed file to get a a plain you know, a a summary of what it's doing, obviously, checking for duplications. You know, from a change management perspective, you know, can we look at, anomalies based on, you know, so Ron just looking at chain changes one version to the next. Can we can we see, you know, what trends we see? What do we see things happening, similar things happening? So, you know, do can we suppress things that we always see and we've always accepted, through to macro analysis, looking at analytics across, IMS, and then, you know, into, some other things around typical to this document summarization, things that people are more familiar with with, AI in terms of natural language queries or assessment of any documents to the solution to see if they are actually meeting, the control statement, in RMS that that they're meant to be, supporting. So that's kind of things that that we've come up with. This as Wayne Belle says, this is not an exhaustive list of things that we've, that we can we can, can put into the technology. We're very keen to hear what what ideas you've got or if any of these ideas have a of any interest. Can we, you know, can we set up working groups to get to the next level of detail of what what actually we do need to to put into product to to make it meaningful and worthwhile and and to drive two things. I mean, AI does two things. It'll drive efficiencies into our programs. I guess, for most people these days, that's table stakes. What we really wanna do is can we actually sort of start generating analysis that you would not normally, be able to achieve without without use the use of AI. So, I'm not seeing anybody coming on stage. So, if you do wanna join this discussion, I can Wayne and I can keep talking, but it'd be better to have a few of you in here. I think I think the idea is you raise your hand, and then the team behind the scenes, Charlotte, will push you up on stage. So please please do come to please do, come and come and, come and join. Or if not, just, you know, ask any any any questions that you may have. You can send them in if you're feeling shy. Just send them in send them in, through the chat feature. As I said in the beginning, the other the other element of, of this is, hi, Brandon. Thanks for joining. The the other thing is, yeah, are you being asked about managing AI? Are are you is your organization looking at managing, managing agents and things? So, certainly a a subject that both, Wayne and I are being increasingly asked about, by clients in terms of, some people are looking at AI agents as EUCs. How can how can we get an inventory of them? So yeah. So for discussion. Anyway, Brandon, thank you for for for joining. What would you like to talk about? Which which bit is interesting? AI in the product or managing AI? Maybe a bit of both. I would say we're going through an operating model review at the moment in terms of segregating what is a combined model and the UC and AI framework into probably something that's a bit more tailored for for the the risks within each. So identifying AI scripts would clearly be a a benefit, but probably most important would be embracing the AI analysis in terms of the data that we already have on our our models and EUCs. So but, I mean, obviously, understanding the the the scripts, so Python scripts and stuff and what they would, anything they're calling, that is existing functionality within the tool. And I guess to your point, you can tag things that are involved in an an AI project. So I don't know if you're familiar with the TSM module, which allows you to do well, discovery will allow you to find all that stuff. TSM will allow you to do the management, of them. Yeah. So at the moment, in terms of the maturity of those processes, we are using discovery solely for for spreadsheets, and and. we're not using TSM to do much with scripts. I'd I'd really love to hear of of any sort of real world successes from from others on the call in terms of actually utilizing all of this functionality. So so I think I mean, the discovery will allow you to to without TSM, will allow you to to look at the scripts as well. Yeah. So, you can and then within shipped within the product, I think there are a number of rules that are identifying AI, suspect not suspect suspected AI, functionality. So things that are calling AI functions, within the Python script. So that that that is that is identifiable currently through the through through the discovery modules, so you can include that in in, and I can't remember if M&G plc has got the the latest pipeline stuff. I can't remember if you're upgrading the latest version, but within the pipelines, you're able to you're able to combine, you're you're able to combine searches of spreadsheets as well as scripts in the same discovery discovery, what do we call it? Pipeline and then the the the same the same scan. So whereas. before we have to separate them out, you can then pull those together. You are then able to so if you if you are trying to do a specific AI search across spreadsheets and Python, that would allow you to do that and start sort of drilling down into that within within the the technology itself. The sorts of discussions we've sort of had beyond that is obviously you can manage the the the, AI the the Python scripts in TSM, but I think a lot of people at the moment is, you know, it is largely taking the information well, creating an inventory of those AI agents that we are creating, putting that into into the inventory, into maybe let's say we've got ease people have got EECs, tools, models, and three different inventories. You can have AI models in a separate inventory. And one thing that we are starting to look at is, is there information that we can get from so for example, if you have a use case in Claude, so you can play maybe create a a higher level object, which is this is Claude. We're using Claude, and then let's say we've got five use cases for Claude. It Claude actually has an API with a whole bunch of information about what it is doing or what people are doing with it. Can we ingest that information through the API, the the IMS API so you can surface that governance information into IMS automatically? So like you have for spreadsheets, you get information from e s m up to up into IMS. For those AI models, claw models, you would be able to get information from claw directly through the APIs. And we we're running a project internally here where, I mean, Mitratech uses we we're we're an AI first, so we've we've probably got we got I think talking to our, chief operating officer, I think we have, we have agreements with all the major AI providers. And so from each of those, we want to be able to monitor what we're doing with that. We have to report to our board and, obviously, up to our investors on the use and making sure that we are within our, risk tolerances as well as, are we utilizing it in the best way? We're not using the most expensive models for simple tasks and things like that. Those are the sorts of things that we are going to be trying to surface through the information that exists, within those models to surface it and trying to surface relevant information. Because there's a plethora of very detailed technical information. How do we take relevant governance information to send this up to people? And that is achievable today, through, you know, the APIs and IMS and what you have. It's just how do we structure it? What does that look like? What what is it? What is and I always work back. What is it you would want to report to your stakeholders? And then how do we how how can how can we surface that from, or how how can how can we deliver that to them through through, the technology you have from And in terms of utilizing the AI on the, you know, the spreadsheet? summarizes, the the macro analyzers, how how embedded are are those tools? So that that that so reiterate what Wayne said. The the issue the issue we have on those is, for if when Alex was talking earlier, a lot of what we can do from an AI perspective because all of the other the the the majority of the solutions that Mitratech provide are SaaS based. So very easier for us to embed those in there. There are, there is optionality or there is there are, enablement switches so you can turn on and off your, your subscription to that. So it it's it's an opt in capability. Whereas, customers on premise. So we need to work on how can we effectively plug into whatever your flavor of your favorite flavor of AI is, whether it's Claude, whether it's Copilot. And then how how can we utilize that to be able to deliver this this type of capability? So we've got an extra step to do, with you guys, in terms of being able to to because then you're in you are in control of your data. You're in control of where you send it to. You've you've got your guardrails in place. So, effectively, you bring you bring you bring your your AI, and we'll plug it in to be able to deliver this type of capability. Is is the high level. But, again, all of these things are relatively easy easy, inverted commas, to deliver. The complication is how do we do it by using you use Claude, someone else's open uses open AI, someone else's new co Copilot. How can we deliver that spreadsheet summarizer or the port portfolio analytics when you all bring slightly different models? And that's what we need to square the circle on. And whether you whether you guys would be happy to send your spreadsheets to Claude and all the others to do that analysis. So we've always it's always been very easy for us in terms of data privacy because we just we we none of your data ever leaves the leaves your organization. It is within your bounds. So that makes it easier probably. Have you have you actually got an adopted AI provider? Have you really got one within your organization? So so we're using, Copilot, I I say, yeah, sporadically. And, you know, in terms of the options within that, so we've got Claude and and another that I that I forget. So so we do have a number. So from our perspective, it's just configuration at that point. So we would basically provide a framework. We would then have to just sort of tap into whichever flavor of AI that you have that you feel most comfortable with, as a solution. So so we've got both GPT and Claude, Okay. available within Copilot at the moment. Cool. Are there any other questions from anybody else? Thank you, Brandon. Wayne Belle, was there anything else you wanted to highlight at all? I know we've got a few minutes left. Not not on this slide. I think you've covered it really well with the sort of AI governance piece, which is definitely a piece of the other piece, the other side of the coin where you we wanted to have the conversation around. So yeah. So I I think for for those who are still on, you know, this is definitely a subject that we want to, to to to drill down into. We definitely want we don't want to develop in a vacuum, so we definitely want to get your input as to where for the top half, where where's the best bang for buck in terms of introducing AI and and then how how we practically can do that when you're bringing your own, AI model to to the party. And then I think the bottom one is how we are being asked to manage AI. Let's say, we we we've got an internal program going on to do that, so we'll once that's up and running, we'll be happy to share that with you. But very keen, to take your input, as we're as we're building that out. So please please don't be shy in in, sending over your ideas, your thoughts, and what you're being asked from the organization because, the more input we get, the more we can build out a a a more rounded solution that meets everybody's requirements. I think, Wayne Belle, if you just go forward to the last slide, just to I think this is a bit of an aide memoir. We do have a number of programs at Mitratech in terms of trying to get, your collaboration. So Wayne Belle mentioned those working groups. We we we want to run those. We're obviously doing these sessions here. We're very happy to do one on one, road map sessions and discussions as well. So, again, reach out to us by your account managers. One other, thing that we are doing is this, customer champions program, CCP program. Effectively, you just sign up. There's nothing you need, to do. If you you can sign up on, using that QR code. Oh, I haven't used a QR code before. Okay. I I think I've I've got an old fashioned link, web web link somewhere as well. But, effectively, what that is is, there's an ability to earn points. So if you are, happy to provide a reference call or do a case study, talk on a webinar and things, then that effectively allows you to to earn points. And of course, as everybody in The UK knows points means prizes. So, there's ways that you can redeem those rewards, and things, you know, everything from a gift card or donation to a charity all the way through to, earning earning earning, money to you can put towards towards services, or we do offer some benchmarking service as well. So there's a there's a number of, benefits that you can you can get from, from, the the hard work you that you put in to to to helping our our customer show jump in program. There are a number of people out there who quite like talking at webinars and things. So if you want to increase your personal profile, again, please, sign up, let us know, what you'd like to to do. Without further ado, I think that's pretty much got us to the top of the hour. Charlotte, do you want to close us out? I'm gonna launch a quick survey for the final two minute. Guys, thank you all so much for joining, and, you know, speaking live with the team. Any other questions, answers, we will answer, in the follow-up, and you will receive this recording by the end of the week. Wayne, Henry, thank you so