Name: Software Monetization Office Hours: Getting Started with Revenera MCP Server
Uploaded: 2026-08-26T15:31:21.614Z
Duration: 39 min 45 s
Description: Software Monetization Office Hours: Getting Started with Revenera MCP Server

Transcript for "Software Monetization Office Hours: Getting Started with Revenera MCP Server": Really excited to be here. Hey, everyone. My name is Ashish. I'm a senior product manager here at Revenera. And today, I would be presenting you, with an introduction about the Revenera MCP server that we have recently launched, in June 2026. So before we go into, MCP servers, AI agents, let's just briefly talk about, what problems, modern businesses are facing. Modern businesses are often required to make faster decisions with increasingly fragmented data. When we deal with entitlement, product, finance, customer success, support, and operations teams all ask slightly different questions about usage, entitlements, renewals, activations, and sometimes even customer behavior. While traditional reporting still matters, it assumes that we already know what the question is, and therefore, we know what the answer would be. And, of course, the the right metric to, to obtain the answer. So it's very easy for us to imagine, ourselves or one of our team members in a situation where they're asking a seemingly short but a complex question. This could be, team members from our sales and renewals team or perhaps even ex executive. And they might ask, what customers renewing next quarter are showing declining usage? The the typical user journey today, for someone using our flex and operations, ALM platform would be that they have to log in to the entitlement management platform, then they would have to, check our reports and dashboards, navigate through those, various dashboards that we provide out of the box, or, perhaps build one of their own. And then from the data that they can see, observe, on the, on the dashboards, they would have to draw their own conclusions and figure out, okay, these are the customers which, which fit the the pattern that we are trying to, uncover over here. But sometimes the problem is, not that simple. And people may ask questions, such as, you know, this dashboard has a lot of information. I don't know what's relevant, to my need, or, you know, I need to check multiple reports to find, the right answers, or I may need to create a fresh report, to uncover the answer, which is not there in the, the existing metrics or the existing reports that we have already created. So the problem, isn't that the customers are lacking data today. The problem is that getting from a business question to a usable answer still can take, a lot of manual effort, and the path from the question to insight is not always that direct. And with the adoption of, artificial intelligence, today, the expectations of customers and users have changed widely. People no longer wish to use traditional dashboards for all their, purposes. Of more often than not, users are more than happy to ask such a seemingly short but a complex question to an AI agent, and they would they would expect the AI agent to interpret what, the customer or the what the user really needs, then identify the data that is needed, query the right system, and return a very clear and comprehensive answer. While that does not make dashboards obsolete completely, and dashboards are still pretty pretty good for known KPIs and repeatable, reporting, and perhaps even for periodic executive reviews, but they are not always the best interaction model available. Something like an MCP server enables this evolutionary step. It allows Revenera and its monetization data to participate in that kind of agent driven analysis and workflows. So, what is, Revenera MCP server? At its simplest level, it's it's a bridge between producers, AI agent, and their own monetization data within the revenue platform. The producer may already have an agent in their own ecosystem, and the MCP server gives, that agent a controlled and a governed way to retrieve the, the data that lives in Revenera ecosystem. The important part is that Revenera is not trying to influence or replace customers' AI strategy. Instead, we have made an attempt to make, data accessible to the AI tools and agents that customers would have already implemented in their own organizations. So this is a a very, common use case or a common model in the industry, which is often called bring your own agent model. On the left, you would see that the producers, or our customers would have their own ecosystems. That would include, a AI models or platforms such as Copilot, Claude, Amazon queue, sometimes even an internal agent platform or other enterprises, enterprise AI tools. On the right, we have revenue ecosystem, which contains, FlexNet operations, CLM, and the modernization data warehouse. Revenera MCP server sits somewhere in between these two while it is still part of Revenera ecosystem. It allows, producers agent to connect to Revenera data without requiring revenue data to provide or manage the agent itself. The that's the clear distinction here. That is important because enterprise customers, many of them especially do not want AI embedded in their existing workflows, perhaps due to security or, other reasons. They may want a separate, they may not want a separate revenue only AI interface. And this, MCP server, allows publishers to manage and maintain their own agents, while Revenera provides a secure data access layer to their monetization data. Let's go briefly into how this end to end flow looks like when an agent or a producer user is actually using, an MCP server, with an AI agent set up on their site. So, it would always start with a question. A a user, a producer user would ask their own, AI, a business question. The the agent would interpret that question and would decide which data, is needed for the insight that the user requires. It then sends a request to Revenera MCP server. The MCP server, acts as a go on access layer. It queries the monetization data warehouse and retrieves the relevant data. That is then returned back to the agent, and the agent processes it into a response that the user can understand. Two points worth emphasizing over here is that AI agent is not reaching directly into the warehouse. MCP server controls that interaction, and, that is a very powerful separation. And second is that agent is responsible for the conversational experience. And when I say conversational experience, that would also include summarizing, reasoning, and helping the user refine the answer. So the governance of AI remains within the producers' ecosystems. So, this may sound, quite exciting, and may seem very promising, but what problems what real world problems, does it really solve? The the value, as I've mentioned over here wait. Just give me one second. Yeah. And, sorry. I I skipped the slide. Okay. But the the value is not just, limited to revenue systems. MCP server are, an an industry wide, platform or an industry wide model, and many other systems, also publish, MCP servers of their own. And the value would then become even stronger when we think beyond a revenue alone. Many customers span the, many customers' questions would span entire, quote to cash process. A single question might need CRM, entitlement, usage, renewal, and ERP information. And, right now, users would have to switch between, different systems to gather, data from these, disjointed systems. However, if you have MCP servers being offered by all of these systems, including Revenera now for your entitlement management, an AI agent could theoretically connect to one single AI agent, would actually connect to all such systems and ask a common question, and the agent would be able to fetch data from all such sources, bring it back to, to itself, reason on it, and provide an an easy, readable insight to the user. Yes. So what problems does it really solve? The the headline would be speed and reach. MCB server helps users find answers to questions that they may not have predefined in a dashboard. It also expands who can ask those questions because the entry point becomes just natural language rather than dashboard specific or, writing SQL queries. So, it it's quite easy to break it down into four specific benefits. The first one is democratization of, analytics. Oftentimes, I I'm asked, internally and externally, you know, who's the the target persona here? Who is this benefiting? And, honestly, I always answer everyone, because not everyone can build dashboards, create reports, or write SQL, but everyone can ask a question and get the answers that they seek through the MCP server. MCP server thus lowers the entry barrier for all the personas. The second is finding answers faster. So traditional way, you know, in order to find, an answer to a question where, insights from the data are required, you have to export data, you have to build a report, and analyze it. With the MCP server coming into the picture, this can be done. What used to take perhaps, maybe several hours or even days to configure to build can be done in a matter of few minutes. Third is exploring unknown questions. Dashboards are excellent for known KPIs and standard reporting. But when the user is investigating a new platform sorry, a new pattern, dashboard may not be enough. An AI agent can help them discover patterns or signals interactively. Something they may may not have known which exists in their data could be uncovered through an AI agent, which is talking to the MCP server. And lastly, connecting multiple systems, together. This is exactly what I had talked about, in my previous slide that, MCP servers allows for an agent to reason across multiple systems, and the overall, message would be that MCP server makes revenue data, easier to access, faster to analyze, and more useful in broader business conversations. Alright. Now that we have upsold, MCP server quite a lot, where does that lead, where does that leave, our existing dashboards, that we have? And it is important for us to, emphasize that it is not a dashboard versus MCP question, where one replaces the other. It is about choosing the right solution for the right type of problem. Dashboards and MCP servers serve completely different needs, and, the next slide should be able to give you sort of a rule of thumb when to use which approach. So for recurring standardized needs, the flex set operations, analytics dashboards remain the the the recommended solution from our side. If you are monitoring KPIs, preparing regular executive reviews, enterprise level report reporting, you have to track trends or, you know, consistent metrics are known to you, dashboards are the right fit there, because they provide structure, repeatability, and a shared source of truth. MCP server becomes valuable, when the need is more exploratory or conversational. And we would suggest that it be used for ad hoc investigations or questions that, do not justify a new dashboard. In in some cases, analysis that involves through, you know, a conversation or follow-up prompts. And, lastly, of course, questions that require reasoning across multiple systems such as CRM, ERP, billing. So, again, MCP server is not a replacement, but an augmentation of your analytics experience. Okay. Now, now that we have understood what the MCP server is and, what it could potentially do, let's have a look at what it really takes to access it, how you can access it, what are the features, and, a demonstration of a couple of, persona based use cases. And for that, I would like to invite on stage my colleague, Ullas. Hey. Thanks, Ashish. That was a great presentation, so I will take our front. Thank you. I'll be waiting for you at package. Okay. Okay. This is how, the producers will actually, the setting up is very simple. Producers will just have to go get the access token credentials either through the FNO application, which is preferred for SSO customers, or through f and o test API. So once this token is generated, you have you can bring your own agent. Either it is a new agent or an existing one that you have. Then you can just simply add a tool, which is MCP, and just add your access token, which was created in the previous steps. Connect it. Once it is successful, now your agent is ready to access the revenue data through MCP Server. It is just simple. So now let's see how exactly or what exactly the data is available in MCB Server and how what are the features or the behavior of the query that happens. As you see, definitely ALM is is directly interacting the monetization data. So monitored whatever the monetization data that can access is available directly from the data warehouse. And the reports that is available, which is, like, account, device, cell device, number of entitlements, all the fulfillment is already available. And for the two calling of your a agents, which is the tools that is available there. Now what are those definitions of the query or some tools that is there? You can only read that data from the data warehouse and you cannot write to it. And the responses will be in the JSON format as a complete record row and without any aggregation. Today, you can actually, by default, date range that you can query per, per query is for twenty four hours as max as three sixty five days, but it is not limited. You can actually refine your prompt in the next, in the next prompt and then get more than the six to five days. And it is also available in all the production environments and in the regular environments. Today, there is no cap for RDD or a daily query, so you can do it as long as you want. Next, I would like to say some of the best practices that we have in setting up your agents. It is just our recommendations. So based on your query, whether it's simple, medium, or complex, you can choose your models. Cloud HICO for simple. If you want for aggregation, it is CloudOpus. Awesome predictive areas, if you want to do complex things, it is SONNET five 4.6. Even the orchestration layer, which is an intelligent controller between URL and the Ravenlo MCT server, is just that you make sure that the decomposition of the complex query is done in retrieval of that subqueries and then summarize all this data. Make sure that you input the compact, not so verbose responses to the LLM so that you get the better reasoning of it. It's just our recommendation. So here is a real deal. I would like to show the some demo of it, and I would like to share the screen. So here I have Hope I'm sharing the screen. So I have set up my Copilot agent, and I have connected my server to this. So I would like to give a scenario. Imagine it's a Monday morning. I am a customer success manager, which is responsible, for dozens of accounts. My challenge isn't getting the data. It is actually turning the data into action. So, traditionally, what I would do is to run entitlement reports, fulfillment reports, utilization reports, export those spreadsheets, correlated data, and then decide where I should spend my time. With NCP server, I can simply have this conversation. So let's say I have these accounts, three accounts, customer account, that I want to simply question it whether my customers are healthy, which one are at risk, and where should I engage them. Me, instead of opening multiple screens and reports, I simply ask a single natural language question. This is where I have in the time interest of time, I have just already prompted it so that I also got a outcome. So as a CSM, I would like to understand how many entitlements have been created and fulfilled over the past three months. And, also, I need utilization summary. If you see here, it is impressive, isn't it? Because it didn't simply return the data. It has gathered the information from multiple sources, correlated the entitlement data with fulfillment data, calculate the utilization, and then presented the information in the business language. If you see an example of subaccount 2026, it is running at 100% utilization. So as a CSM, it it is far more valuable than actually saying just 200 license consumed. What I really care about is, is this customer out of capacity? Is there an expansion opportunity? Should I call them? The agent not just translated this data, but it is also giving me a business signal. Now let's focus take this story which is j r oc two if you see here one of the product is 100100% used which is new embed plot and the rest of them have got 0%. If I look at this reports manually, it would take a significant amount of time understanding what is happening. So instead of that, if you read this, the agent has connected the dots and has given me the narrative. If you notice that this is an upgraded product, which is consumed, which is adopted 100%, but they have never cleaned up the other license footprint. As a CSM, that immediately tells me that this customer might have just migrated to the new product but never cleaned up their license footprint. Now I can have a conversation with my customer as rather than simply asking how are the things going, I can say, hey. I noticed you're getting stronger adoption on the upgraded product, but you still have 100 unused licenses on another entire elements. Can we review whether those licenses are still needed and discuss how we are using the new version? Isn't it much strategic conversation? So and you also see the prototype entitlement. That means that indicates that there is a new capabilities. So, well, it's a both risk and the gross signal in the same conversation. What I love this is that I have not asked for health score. I have not asked for churn analysis or some expansion recommendations. The NCP server has combined the data, reasoning, and the context to generate customer success insights that normally require experience and analysis. So as a CSM, I didn't ask for a report. I just asked for guidance, and that's exactly what Sensitive powered agent deliver. This is one of a kind of persona. Let's switch hats for a moment for a persona, which is technical support engineer. Now, yes, we had looked at the word to the eyes of a CSM. Now just imagine you are a technical support engineer responding to a potential licensing incident. Now here's a scenario. A customer comes and says that, okay. There is an unexpected license transaction, and I want to know who changed it, when it happened, and whether anything unusual occurred. Traditionally, this means opening multiple audit reports, filtering by users, correlating time stamps, and manually building the time limits. Instead, I can simply ask NCP agent to investigate. So if you observe, I have intentionally broadened this prompt. I'm asking Harsh, the user, across the entitlements, line item, device and activation over the last thirty days. And also give me a suspicious activity. Please notice that I'm not asking for a specific report. I'm just asking him to be the agent, to be the investigator. If you observe, the MCP did not just stop at the data source. It has automatically gathered information from entitlement history till the device activations, device history, and line items. Then it is giving me everything in a single timeline. So, basically, instead of me stitching together four different audit reports, agent is just correlation doing a correlation to me. If you see what becomes more interesting is that that out of dozens of audit records across the environment, the agent identified only two direct actions performed with the target users. If you see here, both these actions are at the same time. One has created it, another has deployed it, and Harsh, our user, has modified the entire element. Now instead of hundreds of records, I just have to focus on this story. What happened during that modification? So I'll skip and go to the last what are those flags that has raised by the agent? One is unauthorized full license consumption. That means soon after the entitlement is created, after three hours, Harsh has consumed all the thousand seats in a single action. That's a drastic change, isn't it? And exactly the type of event support teams would spend hours trying to locate it. Next is unauthorized license model change. The license model has changed from this to this. As a support engineer, any licensing model change immediately deserves my attention, isn't it? Because it's a drastic change that may may impact any behavior. And if you see here, the entitlement is created by SK Shiva, but as soon as the update made by Hirsch, it is showing that entitlement is created by Hirsch. Maybe this is expected system behavior or a configuration issue or a workflow problem or something more serious might require investigation. But the important thing is that I would have completely missed it, but the agent highlighted it automatically. So as a support engineer, we re we are rarely asked simply to retrieve it as intent. As a DSP, I'm asked, who changed this? When did it happen? Could this be the root cause? What changed before the customer notice session? Those are investigative questions. What's impressive here is that MCB agent didn't just produce logs. It has produced narrative. It has connected actions, time stamps, users, and system behavior into story that I can immediately use with engineering or any other stakeholders. Please notice that or that I can highlight is that all these results of are the reasoning that is performed on top of the data. The value that is coming from here are combining of three things. One, access to enterprise data to MCP, the domain second, domain specific tools and audit information. Third, the most important one, all the seizing capability of the AI model that you choose. So a different model may identify different patterns, prioritize different anomalies, or provide different recommendations. So here, the MCP as provides access to the truth. The model provides interpretation. So finally, for support teams, the real value is in faster access to logs. The real value is faster understanding of what happened and why it matters. So this is the demo that I will end, and then I'm giving it to back to Ashish for the next steps on the webinar. Welcome back to the stage, Ashish. Thank you, Lars. That was indeed a a very powerful demo, and I hope the audience, would was able to connect to the use cases that you demonstrated and why, it is, a truly an impressive tool that we have, offered as part of our analytics offerings. I I see someone has also asked that, you know, what are the steps to link, a Copilot agent to, the the MCP server, and Anurag, our engineering leader, has provided the the link, in as part of the response, and you can also find the link through the user guide, in the doc sections. Okay. Thank you, Lars. Okay. Let's get on with our final message. So in conclusion, MCB server creates a new way to explore monetization data. It does not remove the need for dashboards. It does not replace customers' AI platform. It connects the two worlds, data and produces agentic workflows or conversational agents. On a broader level, Revenera MCB Server fits in perfectly with the existing portfolio of analytics, of analytics portfolio that we have, and the multiple ways customers can interact with their business data. So dashboards provide ready insights for, recurring business questions for all personas, technical or otherwise. We have, direct database access through Snowflake, which supports customers who want deeper enterprise level data integration. Our, REST APIs, which we call data access APIs, They help publishers integrate our data with, business intelligence tools, or there is existing reporting environments without the need, for logging into, the Flexera platform. And now MCP server adds a new new layer to the portfolio. It's an agentic intelligent cross system analysis from publishers' own environments, and it allows revenue as monetization to be monetization data to be used in conversation analytics and intelligent workflows alongside data from other business systems if the producers choose so. It was really what was needed of us, in these evolving times. Alright. That is all about, the MCP server. Let's have a very, very brief look at what's coming up next in 2026 and a very sneak peek at 2027 road map for analytics. So in q three, we'd be launching customer facing analytics. Customer facing analytics is giving, an analytics experience to our customers' customers in the flex set operations, end user portal. So the end customers would be getting, a set of prebuilt out of box reports that they can view, download, subscribe to, filter on, save bookmarks. And then in q four, we, would be creating, a new set of dashboards, for specific or commonly used user personas who interact with our monetization, analytics dashboards. Make sure that the the visuals on specific, dashboards for these personas are curated for them, and they make lives easier and saves a lot of time for, the people interacting with it. Our '21 our 2027 road map would be, quite AI heavy. We would want to go deep into, predictive analytics. This would be a pilot project, perhaps even a preview feature. We would want to analyze, our customer's data and see what sort of predictive, trends, we could provide to, our publishers and help them with the business needs. And we also have an enhancement to the MCP server plant sometime in, the first half of twenty twenty seven. We have identified some internal improvements, and we are, still hearing great feedback from, customers who are evaluating the MCB server or currently using it. And all of that feedback would be taken together, and, another increment to the MCP server would be delivered sometime in 2027. I would request the audience of, of these officers to please go ahead, go back to, your systems, later today or next few days. Try and set up, an AI agent yourself. And, the the user guide has been provided over here by Anuragha, and it's part of the doc section. It's fairly simple. It'll take you only a few minutes. Once you have that, please ask it a few questions, see how it responds. And whatever feedback you have, constructive or critical, we're happy to hear it and, make sure that it improves the product offerings, going forward. Alright. This concludes our presentation. Before we open up the floor for questions from the audience, we'd like to inform that MCP server is live in all environments for all tenants, external and internal. This capability is, however, available only for those customers who are on the Flexera operations ALM cloud platform, not the legacy ones. Thank you. And, Christine, back to you. Yeah. Thank you so much, Ashish, and thank you, Ulas, for for the great demo. So we'll take a moment here to let folks kind of organize their thoughts. I know there was a lot of content. There's a lot of material, which is why we do this event so that, one, we have it recorded. You guys can watch it on demand later and then come back to any of our team members here with questions, but also to kind of if you have questions now to kind of, allow you to the opportunity to ask the teams live. And I I do wanna reach out to to those folks here. There seem to be some technical issue between the the slide switching. Were folks able to see the slide with the with the road map? If not, I'm happy to put that back up while we wait for folks to, you know, to submit any questions. Let's see here. Remember, silence means you saw it, you're good with it, and then you have no questions about it. So, alright, I'll go ahead and take that as you guys are all okay. Alright. So in the meantime, we do have our, event survey. And for those who might have joined later, like I said, this is just really to ensure, that we're providing you guys events, that are of interest to you guys, of value to you guys. So if there's any feedback of topics you guys would like covered, anything we can do to improve the experience, do let us know. And then there is also, as Ashish had mentioned, the docs tab next to the chat, which has the user guide. There is a a survey the team would love you guys to take. Sorry. There's a bit survey heavy, but that would help us with the kind of future considerations for our analytics solution as well as the MCP server. So those your feedback is valuable. And then last but not least, if you've not registered for the upcoming user group, which is happening in October, please do so, because we would love to have, have all of you guys there. Alright. Looks like no further questions here. Why don't we go ahead and conclude the event then? Again, Ashish, Ullas, and Anuradha who is lurking in the background, so don't think we we bait switched anything here. She is here. You know, they are here to answer your guys' questions about the MCP server, should you have any in the future. And thank you again for attending. Hope you guys have the rest of your day, and give you guys fifteen minutes back. Look out for that recording. Thanks all. everyone a good. one.