149. How BMC’s CCO Built AI Agents That Triple Productivity ft. Sofia Barbosa (BMC Software)

It’s much better for me if I have my CSM on a quarterly business review, visiting a customer, working on adoption, or having much more strategic relationships and actually increasing connections we have inside the customers to other departments than fulfilling documentation, writing the stories. They are already inputting a lot of the data we need. I’m just giving them support and I think that will make their role much more interesting.

Josh Schachter [Host]:
You’re listening to Unchurned, brought to you by the Gainsight podcast network. Last September, when BMC split into two companies, Sofia Barbosa was promoted to Chief Customer Officer. Smaller team, same massive challenge, supporting hundreds of products globally. But here’s what happened next. Within weeks, her team tripled their output of customer success stories. Processes that took three weeks now took hours. Support escalations dropped to zero red alerts for the year.

Josh Schachter [Host]:
How? Sophia didn’t hire more people. She didn’t reorganize. Instead, she built an AI platform called Succedo. What used to take months creating a customer success story now takes a day. Support tickets that might escalate. AI catches them first, trading paths through hundreds of courses, personalized instantly. But this isn’t about replacing people with AI. It’s about Sophia discovering that when you free CSMs from administrative work, something unexpected happens.

Josh Schachter [Host]:
They can finally do the strategic work that actually moves the needle. Hey everybody, and welcome to this episode of Unchurned. I’m Josh Schachter, your host, and I’m here with the CCO of BMC, Sofia Barbosa based in Portugal. Sophia, thank you so much for joining us.

Sofia Barbosa :
Yeah, thank you so much for inviting me, Josh. Really glad to be here.

Josh Schachter [Host]:
So we all know BMC. I mean, major enterprise. You guys do so many different things. Big company. Tell us very briefly what BMC does and then I’d love to learn quickly about your org.

Sofia Barbosa :
Yeah, for sure. So BMC is a leader on the mainframe modernization space through our Amy suite products and also in terms of workload, automation and orchestration through our Control M product. So we have customers all around the globe. And my team in particular in customer success is divided in several different types of areas, like professional services. So we have a team of experts that supports all of our products and implements solutions. We have our customer success management and customer success specialist group that basically helps customers with value realization, with adoption and all of the good things we want our customers to achieve. And we also have. I also have the support organization under my Remit and Customer Experience, which handles with all voice of the customer and all related activities in terms of customer experience and outcomes.

Sofia Barbosa :
So in a nutshell, we are a global team. We are across all continents and it would be in excess of 415 people at this point in time.

Josh Schachter [Host]:
Wonderful. And you were not so long ago promoted to the role of Chief Customer Officer. There was an organizational transformation, if you will, a little split in the company and some. So now you have BMC and BMC Helix, and so you’re running the BMC side of the org. I don’t know if that’s the official title for your side of the org, but congratulations on that. What went through your mind when you got the nod that you were going to become the cco?

Sofia Barbosa :
Yeah, so I think so. BMC did announce a company split back in September last year and we had to readjust all the organization to be able to support these product lines and the company. I was already managing support in customer success management previously. So I think it’s an interesting challenge. I think the company is growing at a fast pace and I’m really keen to see where I can take customer success forward. And I think sometimes there are things that moves us into a certain direction. And I think one of the things that I really want to share with you, Josh, and with everyone today is how this split made me think even a bit deeper in terms of how CS is operating and how can we automate more, how can we be more efficient, how can I ensure that my customer success specialists and managers are actually spending more time with customers and what can I do to help their day to day? So I think it was like propelling us forward for scale automation and how can we help our practitioners in the field? A consequence of this fleet and the way we were organizing the team.

Josh Schachter [Host]:
Okay, so I want to get into the deeper thinking, how it forced you to think deeper about the org. But first, yes, I’m friends with Colin Murphy. You know Colin?

Sofia Barbosa :
Colin really well, Very well.

Josh Schachter [Host]:
You worked for him. Former cco, bmc, now CCO at Zendesk. And I emailed Colin, I actually texted with him and he sent me a voice memo. And I asked Colin, tell me about Sophia specifically. I want to know when Sophia, when is Sophia at her best? Where does she spike? So you want to hear what Colin had to say?

Sofia Barbosa :
No, I won’t. I’m just like, no, Colin wants.

Josh Schachter [Host]:
No, it was very complimentary. Sophia is a woman of many talents, but I’d group her strengths into two categories that together make her unique. So Category one, his words, solving complex customer issues. She’s exceptional at jumping in to untangle complicated customer related challenges, often involving multiple teams across an ecosystem. She doesn’t just put out fires, she puts structure behind the solution, especially when the problem spans more than one customer. One B for him is structuring ambiguous problems. On the other end of the spectrum, she’s equally strong at bringing clarity to undefined opportunities. Whether it’s launching a new monetized CX offering, designing a new coverage model, shaping a competitive play, or rethinking digital engagement.

Josh Schachter [Host]:
She can take a vague idea, structure to a plan and make it actionable.

Sofia Barbosa :
What do you think, Colin? Yeah, I think he knows me well. We work together. I really like Colin. I think Colin is a great person to work for. I think he’s doing great now in Zendesk and he knows I wish him all the best. And I think we link a lot on the part of the transformation, resolving problems and also transformation and how can we propel an organization forward. So really thankful for his insights. Thanks for sharing, Josh.

Josh Schachter [Host]:
Yeah, yeah, He’s a great guy, that one.

Sofia Barbosa :
That was great. Yeah.

Josh Schachter [Host]:
Okay. So. So, so I think you mentioned that you share a penchant for transformation and insights. So I reached out to you. I’ve known you, but I specifically reached out to you a couple weeks ago. You had posted about this thing called Succeedo on LinkedIn. It was very compelling when I saw it because it was all about how you’re going through AI transformation. And it seemed to me that you had very discreet wins that you were sharing, that you had already been, in my perception, a step ahead of many other leaders at an enterprise of the size and scale of bmc.

Josh Schachter [Host]:
So that’s what we want to talk about. We want to talk about tsuxedo and about your AI initiatives, your transformation initiatives so far at bmc. Since you’ve taken this post, tell us at high level, what is Tsuxedo? And then we can go from there.

Sofia Barbosa :
Yeah, so tsuxedo is basically ki solution. We started experimenting. The idea here again is when we did a split of the company, we had to divide a couple of teams in terms of the technical expertise and also the roles. And we needed to move faster, we needed to have a bigger reach. And therefore we decided that we really needed to bet on a couple of tools to make us more efficient. And tsuxedo was one of them. So we started experimenting with Copilot Studio in disguise. But we can use several other platforms.

Josh Schachter [Host]:
But this is Microsoft.

Sofia Barbosa :
This is Microsoft. And the first thing we did is we wanted to ensure that we were creating success stories for specific use cases. And we had the target and we wanted to develop it further to help other customers see how other customers are actually implementing our products and could be of benefit to them. And our CSMs are great in many, many things, but some of them are not writers. So they needed some help. And we decided, okay, let’s try to create an agent that goes and fetch data from these different data sources that we have and actually helps consolidate and put the story together and then we refine it as we go. So we started working on that. We linked to the data sources.

Sofia Barbosa :
We have Salesforce, Gainsight, many others. And then we started creating Succeedo for success stories, which is the practitioner can put the name of the company and the use case they want to put forward and Succeedo proposes and writes the success story. And then we validate and we implement in many cases we send to our advocacy team and to others. So the funny part was through that experiment we were able to triple the amount of success stories we were creating. So we were excited with the results and we thought, God, this is working. So we haven’t enormous list that can follow this one. And then we define a type of a backlog.

Josh Schachter [Host]:
So, so you’ve got your backlog, we’re going to go into that. You. You saw a 3x increase in the case studies, the value stories that you’ve been creating.

Sofia Barbosa :
Yes.

Josh Schachter [Host]:
How did you, and sorry if I missed it, but how did you come to start with creating the, the value stories as your first go into AI?

Sofia Barbosa :
So that was the first one because there was a particular outcome we wanted to achieve with the number of customers. Some stories we wanted to create on a specific use case for a product. And if I was going to ask our practitioners to do that, I would know that that would take me maybe one, two months just to be able to gather all the information we needed and to write it with the level of quality that we think we need to be able to push it forward internally and even externally. That’s why we decided to think, okay, let’s try to create some automation on this and see how can we help our practitioners and have results faster. That’s the way the genesis of CIXIDO started. And in a couple of weeks, I don’t think even if it was three weeks, we were able to triple the number of success stories for those particular use cases. So that sparkled this initial movement, that this works, this brings results and we want to continue to experiment with other areas.

Josh Schachter [Host]:
I think you Said it was taking one to two months to create these case studies. Uh, how long does it take currently with Tuxedo?

Sofia Barbosa :
Yeah, with Tuxedo you can. If I ask for a specific use case, it could take me one day, depending on the volume I want.

Josh Schachter [Host]:
It’s very, very fast order orders of magnitude faster. And is there a human in the loop part of this process as well?

Sofia Barbosa :
It is. So there is always a validation, so six produces the story, but there is always a CSM that looks at the story, validates that everything is correct, and then ensures that is published or internally or to our advocacy team. So there is always a human in the loop. Tsuxedo is speeding us up as you can manses in terms of timelines, but there is always a human in the loop.

Josh Schachter [Host]:
So you’ve got 3x the number of stories. How does that impact the business both within your org and even cross functionally.

Sofia Barbosa :
A lot, because we were able to send much more and explain what our customers were doing in our product, for example, to our advocacy team. And from that we got much more references and we were able to incentivize more peer reviews. So everything that is linked with references, case studies, peer reviews, I think it was a consequence of us speeding up the success stories with Tsuxedo.

Josh Schachter [Host]:
I don’t know if you call them CSQLs, but effectively that. Right, effectively you’re helping to bring leads to the party.

Sofia Barbosa :
Yeah, csqls might be a consequence. This for us is really how can we have targets that are linked to the marketing team, for example, in terms of the number of references that we want to have outside the number of advocacy events. So I would link this more to that area. Although at the end of the day, the use cases development does spark the possibility of creating CSQLs qualified leads here.

Josh Schachter [Host]:
Have you launched this at full scale in your org yet?

Sofia Barbosa :
Yes, it’s fully deployed. Tsuxedo is fully deployed for value for these customer stories, but we are now introducing new ones for different things. Some of them are still in development, some of them are in early poc.

Josh Schachter [Host]:
Okay, great, let’s talk about them. What’s next? What’s next in the pipeline?

Sofia Barbosa :
So the next one for me is value summaries, what we call here internally value summaries. What is this about? Value summaries is we want to ensure that we are able to give to all of our customers, at least as an ambition, once per year, a complete narrative of the value they are retrieving from our products. Of course, it has for Genesis, the outcomes we create the customer Success plans that we input into Gainsight and others. And what this agent does is looks at even from a pre sales perspective. What was the initial business case we had proposed? The customer, what have we achieved through the adoption? Progression to everything we put in Gainsight in terms of progression of the work with the customer. And then it provides you a document that explains the customer, their business unit, how they are using our products and what are the value and benefits they are taking from it. So again there is always a human also in the loop. We are pocing this part, but it simplifies the process a lot.

Sofia Barbosa :
So you reduce it significantly the time we would take to do this one by one.

Josh Schachter [Host]:
And what’s the end business impact of.

Sofia Barbosa :
Having more value summaries we can articulate? What is the value that we are bringing? What is this in terms, even from a metrics perspective, what is the impact in efficiency in our customers, in productivity, in transformation depending on the business outcomes they want to achieve. So we are much better equipped for a discussion in terms of growth or a renewal discussion, for example, because we have this, we document it, we discuss with the customer, we get their sign off and then it’s easier to understand what are the benefits that customers are getting from our solutions or if they are not getting, what is the action plan that we need to put in place to ensure that they are able to retrieve what they were expecting or eventually even more.

Josh Schachter [Host]:
And is it continually updating? I assume that’s part of the magic here is that you don’t have to manually do it and keep it updated. It’s doing that on its own.

Sofia Barbosa :
Yes.

Josh Schachter [Host]:
Yeah. And where does it live, those value summaries once they’re created?

Sofia Barbosa :
So it’s actually we generate a PDF document that we discuss with the customer, but it retrieves data from Salesforce but use our CRM retrieves data from Gainsight and others and through Tsuxedo actually Tuxedo even also looks at data that is available online. If there was something that is significant on the customer story that happened, it actually puts us part of the narrative in the beginning explaining the customer business. So it does help provide all of that information and then we produce the PDF that our CSMs discuss on a KBR with the customer or ensures that the information that is there is correct and is acknowledged.

Josh Schachter [Host]:
So we got customer stories, value stories, value summary. There’s a few more. Tell us about the next one.

Sofia Barbosa :
There’s a few more. Another one is we have a big suite of products, let’s put it like that. I’ve spoke in two categories. But we have more than 400 products overall inside these categories that we are supporting. So we have a big training courses. So we have lots of training courses, certifications and sometimes can be a bit complex to navigate all the training offerings that we have. So we created tsuxedo for Education, which basically is in internal use right now. I’m really keen to push this to our customers as a next step, but it searches everything we have and produces a training path for you as a user or for your team if you’re an admin.

Sofia Barbosa :
So the idea is really that it’s able to help you navigate all of our training documentation, videos and everything we have available and simplify it in a manner that gives you like a prescriptive path just for you. And this for us is very powerful because as I said, it can be complex with the number of courses that we have in learning paths.

Josh Schachter [Host]:
What’s the overall goal of this simplified personalized search? Is it adoption? Is it better onboarding Is adoption is.

Sofia Barbosa :
Better Onboarding is ensuring that our customers are using their time wisely from a training perspective and they are getting into the context that they need at the time they need it. So it does help a lot with speed and does help with onboarding, does help with adoption and does help us ensure that if there is anyone new to use our tools, they have the assistance they need to be effective on the learning they are choosing to start with and then to navigate through.

Josh Schachter [Host]:
Yeah. And is there a way that you can quantify the ROI on this particular agent?

Sofia Barbosa :
Not at this point in time because we are still using it internally. But I do think that is going to be messy for us. We just have.

Josh Schachter [Host]:
How will you quantify that?

Sofia Barbosa :
So first we just launched a customer portal, also a new one, and I want to deploy then succeed the free education. There we will qualify first I think we need to track the usage afterwards is through training attendance. So what are our customers doing in terms of training attendance? We also have surveys for training, how are they scoring? And I really want to know more feedback about Tseedo, so maybe I can even try to get some scores about the help that Tseedo gave until they reach the learning content. So I think it’s usage, it’s completion of training, it’s actually then reflected on the adoption of our product also for our customers. So there’s lots of things that I think I can create a metric upon to check the impacts of succeeded free education.

Josh Schachter [Host]:
Okay, we have a couple more. Couple more agents to talk about.

Sofia Barbosa :
Yes. So another one and this is more on the, on the support side. So our support team deals with all the inquirers from customers and very close contact every day. So here we have another tool in this case is analyzing customer sentiment and ensure that helps us create prioritization of the backlog we are dealing with.

Josh Schachter [Host]:
You can say, by the way, Sophia, you can say the names are other tools. It’s okay. Support Logic, we want to. We’re equal opportunity promotion here. Yeah, yeah.

Sofia Barbosa :
Support logic is the tool that we are using. Basically what it does is it looks at all of our backlog of the tickets or the incidents that we are dealing with and does some prioritization in terms of the sentiments. So reads the sentiment of the customer and say, okay, this is about to escalate. You should take care of this one now. So you can have, for example, on the support side you have usually P2s, P1s, P3s. P3s are the most normal one. It’s a normal incident that you usually depending on the company, you have two, three, five days to resolve. But inside those P3s, maybe there’s a P3 that is critical for a customer and the customer is really nervous and he needs to get it resolved.

Sofia Barbosa :
And what SupportLogic does is analyzes that customer sentiment and says okay, on top of your backlog, this one is very critical for this customer. So please ensure that you deal with this whilst you’re dealing with the others inside the SLA levels. But this one is critical so it helps us with backlog prioritization.

Josh Schachter [Host]:
Kind of reminds me of the Eisenhower matrix. Are you familiar with the Eisenhower matrix? If something can be, you know, the level of importance and then also the level of urgency that two by two it feels like. So, so is that kind of what it’s doing? Is it adding that additional layer or is it actually stepping in and saying no, no, no, you marked this wrong. This actually should be lower priority than what you’ve called it.

Sofia Barbosa :
No, it’s just raising the flag in terms of the ones that we think that it is really critical that we push forward. So it just has a type of amber, green, red in terms of sentiment and ensures that our managers and our support technicians are able to take with the ones that are more press things for customers first. So it does help with prioritization. Where have I seen results in metrics on that, on those ones, on the number of escalations that reduced, on the number of crit sits, on the number of alerts. So it does help us to act faster and prevent escalations from happening.

Josh Schachter [Host]:
That’s awesome. That’s great. Okay, I think you have one more.

Sofia Barbosa :
Yes, we have a lot in the backlog also. Let me tell you something that I.

Josh Schachter [Host]:
Okay, what’s the most exciting, the biggest one that you want to talk about?

Sofia Barbosa :
This can be very exciting to me. One of them is statements of work. And this is because inside CS we also have the professional services, organization and statements of work. Many times when we are doing a big project and we need to document everything in terms of the deliverables, what we need to do with phase one, phase two, what are the dependencies on our side, on the customer side, if there is another partner involved. So it can be quite timely to do a type of statement of work in complex engagements. And we are producing now tsuxedo for statements of work which basically will have again access to all of our data. So it will have access to customer records on our side. And I do think that this will help us simplify and provide us a pre populated template of statement of work in a record time.

Sofia Barbosa :
And then we just need to deal with some things that might be specific. But I think this has the potential to save us per each one around three weeks manual work.

Josh Schachter [Host]:
Three weeks of manual work. Probably a lot of cross a lot of collaboration in there between different parties that are trying to assemble data. And you’re basically. You basically have given the system some examples of your best templates of sows. It understands information about the specific deal and the customer and then it recommends the template that’s best fit and fills.

Sofia Barbosa :
It out for you completely and analyzes customers that have had similar scenarios and proposes a solution based on similar scenarios that then of course a human being or an architect needs to go in review. But it does a big portion of the initial work of design, let’s put it like that.

Josh Schachter [Host]:
Are you going to try to have it close the loop and go even further in the interaction with the customer? I mean, are you going to get to the place here where it’s sending out, you know, DocuSigns or whatever you’re using for signatures, the envelopes and integrating with the CPQs and all those sorts of things?

Sofia Barbosa :
It’s funny you say that because I was discussing with my team today and on the value summaries we are now automatically creating the PPT or the PDF. So what we want is that automated creation and then if we already revise, if everything is okay, it sends to customers automatically, particularly on the lower segments where we do not need that much detail. Is this more simpler usage of Our product.

Josh Schachter [Host]:
So there could be a world where it’s automating the entire workflow.

Sofia Barbosa :
That’s what we are now investigating on this case. And if this can be done there then it could be done for segments of work or other others.

Josh Schachter [Host]:
That’s awesome. That’s. That’s amazing.

Sofia Barbosa :
That’s very exciting.

Josh Schachter [Host]:
Yeah. I mean that’s the future of true agentic. Right? I think there’s this Agentic is an evolution. It’s a spectrum. A lot of the things that that are agentic are kind of an extension of gen AI of being able to prompt and put really good spit and polish and intelligence on top of the information. But in the case of what I think is like the next it’s current but it’s also, you know, the forefront of agentic is being able to take those actions and close the loop and go end to end. And it sounds like that’s one that there’s a real big opportunity for.

Sofia Barbosa :
Completely. Yeah. Agreed.

Josh Schachter [Host]:
How do you manage all this Sophia? What else do you do in your day job? I mean how do you. How did. How did you get this Seems like such a large lift to get off the ground. So what was your process? I mean I guess you started small. You said you started with the value stories, the customer stories. Sorry, I forget which what the official name you called them were. But yeah. How do you manage all this?

Sofia Barbosa :
I’m passionate for innovation so this is really like what excites me because I think there is a lot we can do. We started small, we saw results and all my team is not just me so I have an amazing team behind me that is really excited with the results and we are looking at the market with others are doing. We are trying to. To speak and understand and we do think that in some areas we were able to find some good results already. So we really want to push forward. The conversations with my team recently is we are actually channeling here roles in a different direction because of the support we need to give to the agent. And I know since the last time we spoke Josh, we already have names for the team so we are going to have the developers of the strategy for Tsixedo is our six Forge team and then the people that will maintain and support the agents is going to be the Tuxedo Care. So we have the Tuxedo Forge and the Succeedo Care.

Sofia Barbosa :
I think if you’re so just to time out there.

Josh Schachter [Host]:
So you’ve created entirely new teams in your orgs for to manage the agents?

Sofia Barbosa :
Yes, yes. So we are. We just Reclassified some roles to be able to have people dedicated just to do the strategy and the initial creation of the agents and then the testing, the validation and the ongoing support and maintenance to TSUXEDO Care. So we have the TSUXEDO Forge and the TSUXEDO Care, which is the team that creates and then the team that supports the agents once it goes live.

Josh Schachter [Host]:
And what’s the sentiment around the organization from your team? I mean, are people, people running into the corners not wanting to get picked for this? Or are they stacking up their applications and putting their names in the hat? Or is there a little fear as well? What are people thinking?

Sofia Barbosa :
I don’t think there is fear. I think there is a bit like this is a change management. There is a bit. Okay, let me see what this will do for me. Maybe a bit suspicious in the beginning, but once I think they started seeing the results on value Stories, they are coming up with new requests, so they are increasing the backlog.

Josh Schachter [Host]:
So they’re excited.

Sofia Barbosa :
They are excited.

Josh Schachter [Host]:
You know, the change management piece is interesting. How are you managing the change management?

Sofia Barbosa :
Yeah, one of the things that I think is important as a leader in this area is particularly with all the hype around AI. And my true objective is I’m not planning to replace people with AI. I’m planning to put AI supporting people. And that’s truly the objective here because it’s much better for me if I have my CSM on a quarterly business review, visiting a customer, working on adoption, or having much more strategic relationships and actually increasing the connections we have inside the customers to other departments then fulfilling documentation, writing the stories. They are already inputting a lot of the data we need. I’m just giving them support and I think that will make of their role much more interesting because they can focus on the value creation of the role and not so much on the what we can call more admin tasks that they needed to deal with.

Josh Schachter [Host]:
Have you carved out a specific role for change management of AI?

Sofia Barbosa :
I have not. We are doing this through the central team. We have a central team that helps manage all our main projects and initiatives and they are the ones bringing this. We are doing enablement training. We have regular team meetings explaining. So we are navigating the change and we are explaining what are the benefits. And so far I’m not seeing my team pushing back. I think they appreciate the help is what I’m seeing in terms of feedback.

Sofia Barbosa :
But they needed enablement in training and they need to understand how to operate and how this will benefit them.

Josh Schachter [Host]:
Last question. What’s a hot tip you would give to another CCO at a similar type of enterprise around AI transformation.

Sofia Barbosa :
I think there’s lots of discussion about AI. What is the return on investment, what is not? I think CFOs are starting to ask a lot from CCOs and other heads of the departments you invested here, there. What is the return you’re having? So what are you able to show me? Show me metrics, show me results. And I think we are all trying to get something, but it’s very, very early days. So my advice will be start experimenting, because with experimentation, it brings results many times and brings clarity into what can we expect in terms of gaining more productivity or more efficiency? We have done sensitivity matrixes already on that based on the initial results we had. So I think we need to stop sometimes speaking so much, and I think we need more action. With action will come knowledge. Yeah.

Sofia Barbosa :
And experiment and fail and fail. It’s normal. We need to fail and then recover from it and learn and move forward as with everything.

Josh Schachter [Host]:
Fail fast, fail forward, as they say.

Sofia Barbosa :
That’s it. Yeah.

Josh Schachter [Host]:
Sophia, tremendous work. Congratulations. I know you’re still at the starting line, really, in the grand scheme of things, but congratulations on all the strides and accomplishments so far with your AI transformation. And I’m very much looking forward to hearing about how this all goes and what’s next.

Sofia Barbosa :
Okay, thank you so much, Josh, and thank you for having me.