The Deep View: Conversations

Jason Hiner (00:02.694)
All right, well Shabani, why don't we start, you talk to us a little bit about what your role is at Salesforce.

Shibani Ahuja (she/hers) (00:09.208)
Sure, I've got a, I think, the coolest role at Salesforce. But before I tell you what the role is, maybe I'll give you bit of my past, the present, the future of where, how did I end up here? Because it's not a common role. Title is SVP Enterprise IT Strategy. You tell me if you can figure out what that means.

Jason Hiner (00:17.232)
Sure.

Perfect. Okay.

Jason Hiner (00:25.296)
Ha ha ha.

Shibani Ahuja (she/hers) (00:26.734)
It all started with Salesforce is notorious for having customers come and tell their stories, like their customer success stories. about a year and two months ago, I was working at one of the largest banks in Canada.

And in that organization, I'd done a few things, one of which was I ran digital for the organization, Digital Channels. And my last role was I was the owner of all of our shopping technology, so all the technologies that enabled a customer shopping experience. Long story short, we implemented a, we had a very, very great Salesforce implementation to drive personalized experiences at scale for a customer and their channel of choice.

I joined Salesforce in one of their internal employee meetings to share the customer success story. And there was a question that was asked at the end to say, Shivani, you've got Barq and you've got the top leaders of the organization sitting right here. What advice do you have for them?

And I thought very carefully, should it be filtered or unfiltered? And I think I went the unfiltered route, unfiltered, I let it rip. And I said, look, this is what the experience is as a customer of yours. It's ironic you're helping me put my customer at the center in my bank. But yet, when I show up, I don't feel like you're putting me at the center. I have to learn your products. I have to learn your construct. So it on and on and on.

Jason Hiner (01:25.488)
Ha ha ha.

Unfiltered? Good.

Shibani Ahuja (she/hers) (01:48.158)
It was probably love at first sight. In the moment that I was asked that question, in the few moments that I had spent time with the leadership team prior to presenting, in the moments thereafter and the rich conversation that took place of people saying you really do understand the context of the customer, in this case it was enterprise CIOs, it was a mutual fit to say this is the kind of organization I want to be at and they had said we'd love to have someone like you that has experience feeling what it's like to do business with us and help us make that better.

Jason Hiner (02:15.886)
Hmm.

Shibani Ahuja (she/hers) (02:18.212)
if I fast forward to what I've been brought on to do is I'm leading the organization's go-to-market strategy for how we engage the C-suite of enterprise organizations, so largest organizations. We're in a very different space and place now where...

We are so much more than just a CRM company. And you're probably hearing a lot about that in the news to say we are the number one, this aspiration to be the number one agentic platform for organizations that are heading up on their agentic journey. And so what that's happened, what's happened is I've, in the last year, I've met with over 587, and I'm adding to that counter now, C-suite leaders as essentially a type of transformation and AI advisor.

Jason Hiner (02:55.365)
Okay.

Shibani Ahuja (she/hers) (02:59.598)
And in many cases last year, it was just basic education on what agentic was in a definition and in a way in which people would really understand it. And so it is actually meeting on the ground with leaders that are trying to figure this out as fast as their organizations are moving, as fast as MIT studies are being published on the success or lack thereof, and working with them in listening to what are they experiencing, what are they exposed to, and how can we help.

Jason Hiner (03:28.905)
587, did you say? Like that's more days than there are in, you know, that you've worked there. That's incredible.

Shibani Ahuja (she/hers) (03:33.23)
You do exactly. You tell me. I achieved my airline status in six and a half months.

Jason Hiner (03:41.065)
my gosh, my gosh, incredible. how do you spend your days, how do you spend your weeks? With that, you're going and meeting clients where they are. You're talking to them about their journey with the products, but also I'm sure they're often, the folks you're talking to are trying to think 12, 24, 36 months out, right? So you're working with them on not just what they're doing and how their experience is, but where they're going, yeah?

Shibani Ahuja (she/hers) (04:10.957)
I think what you're describing is more of what I would expect this year to happen. To be very frank about it, last year was spent almost doing a bit of agentic AI 101. That's really like it is, and it's not a surprise when you think about the largest organizations like Fortune 200, the C-suite leaders of those organizations, have to be on top of so much.

Jason Hiner (04:15.416)
Okay.

Jason Hiner (04:23.522)
Okay.

Shibani Ahuja (she/hers) (04:34.976)
And suddenly comes this emerging technology. AI is not emerging, but agentic AI? I think it was a term that was just birthed a little over a year ago. And CIOs or C-suite leaders are all trying to figure out what, well, what is it? Because generative AI made a pretty big scene. How is this agentic thing different than generative and different than the chatbot that we would have implemented once upon a time? And so much of the time was not even spent looking forward, future looking. It was more of,

Jason Hiner (04:42.937)
Yeah.

Shibani Ahuja (she/hers) (05:03.546)
What is this? How do I digest it? What do I need to understand? And how do I have to think about how we're gonna modernize our architecture once they understood what it was? And it was profound to see the shift in maturity and therefore the thinking and how particularly CIOs wanted to architect stacks. A lot of patterns that I saw over the course of the year. Yeah.

Jason Hiner (05:21.305)
That makes a lot of sense.

We're going to talk a lot more about that and about where that goes and a genetic AI because there's so much to unpack there, so much that you all are doing and so many ways you're helping, I think, enterprises digest this and prepare to do it. But before we do that, there's one other question I want to ask you because I don't know where you find the time, but in addition to all of those meetings, you also do podcasts, you speak at events, and I've come across your

Shibani Ahuja (she/hers) (05:46.584)
Yeah, I do.

Jason Hiner (05:52.583)
in that way. You've spoken to members of my team before. You are...

Shibani Ahuja (she/hers) (05:56.044)
Yes.

Jason Hiner (05:59.525)
an exceptional communicator of these kinds of things. In the tech space, it's sometimes missing. We easily fall into both some tropes as well as jargon and acronyms and things like this. And it can very easily get undecipherable and sound like the same sort of talk track over and over again. You have the ability to, I think, elevate things to first principles

Shibani Ahuja (she/hers) (06:18.402)
Yes.

Jason Hiner (06:29.479)
to what are the core ways that you can understand this and that is understandable beyond just a pure technologist, you know, as well. Obviously, you're speaking to a lot of business leaders and I want to know, like, how did you get there? Like, what is your history as a communicator because you truly are so good at it and one of the best that I've seen in the technology space today because there's a lot of what I would call, I'll be generous and say, mediocre communications, you know, in tech today.

Shibani Ahuja (she/hers) (06:41.73)
laughs

Shibani Ahuja (she/hers) (06:47.933)
man.

Shibani Ahuja (she/hers) (06:59.534)
I, you're making me blush. That's gonna be the kindest thing that I've ever heard anyone say that's really sweet. you know, when you're asked to be a speaker on behalf of an organization, you can receive a script and you can memorize it. Or you can actually believe in what the company is doing. You can actually believe in what the customers are delivering. You can actually believe in what the technology is doing. But true belief comes in understanding it yourself. And so,

Jason Hiner (07:01.952)
Hahaha.

Shibani Ahuja (she/hers) (07:27.948)
before I prepare for anything, whether it's a keynote where I'm given a script, it's like I'll say like, I don't get it.

Can you make it real? Can you make it real? And I break it down just for myself. Yeah, I've done a lot of public speaking, which I never did in my last role. So this is all a year old in discovering. Yeah, this is all discovering over the course of the year. But the way I get comfortable in front of a crowd, whether it's you in this intimate conversation with many friends that'll join afterwards or whether it's on a stage, I really just want to understand and take away the technical jargon, take away the consulting jargon, take away the...

Jason Hiner (07:35.46)
Mm.

Jason Hiner (07:43.651)
Really? Okay.

Shibani Ahuja (she/hers) (08:03.896)
marketing jargon, help me understand from a persona point of view. Help me understand from the end user. Is that a customer? Is that a colleague? Is that the administrator? Is it the CEO? So oftentimes when I'm using examples, I really just try to break it down to say who is the human that is benefiting or working with if it's an agent, if it's AI, because that's how we really understand and relate. And I also like to use analogies. And sometimes my analogies are, I call them four-year-old analogies because I have a four-year-old son.

And sometimes when you like take yourself out of the space of the area or the industry or the persona that you're in, it makes it translatable for everyone. And that's always my objective is if I can understand it, hopefully I can relate it, relay it back that way.

Jason Hiner (08:33.413)
Nice.

Jason Hiner (08:49.136)
that.

It's fantastic advice. So listeners, if you get nothing else out of this podcast, take that note, use it when you have to communicate things, and especially very technical things to audiences. Know your audience and break it down. right. Well, Shivani, thank you so much for that. One of the things you talk about a lot, of course, is Salesforce vision of the AI-driven enterprise. You spent a lot of time doing it over the past year, as you said, breaking it down.

Shibani Ahuja (she/hers) (08:53.102)
Yeah.

Shibani Ahuja (she/hers) (09:13.346)
Yes.

Jason Hiner (09:19.431)
I wonder if we could go back a little bit there too and talk about the trajectory of Salesforce's AI vision, going back to earlier iterations of things like Einstein and CRM integrated kind of co-pilots to today because it's quite different as I understand it. It is.

And as I digest it, it's very different from what I think of as sort of the Einstein CRM era.

Shibani Ahuja (she/hers) (09:50.868)
You are so beautifully teeing me up. It was as if you planned that. I'll answer that with probably one of the hottest topics that I talked about last year.

Jason Hiner (09:54.533)
Amazing.

Shibani Ahuja (she/hers) (10:03.094)
which was AI is the smallest, biggest word, or initials. And why I would say that is because you'd be sitting in a room with technologists, CIOs, and their teams, and our sales teams, and I would just, I would be like a tennis match, I'm just listening to them, and they're all using the term AI, AI, AI.

Jason Hiner (10:08.313)
Hmm.

Shibani Ahuja (she/hers) (10:22.814)
And if you listen really carefully to the words that they're using, one of them is describing predictive AI. One of them is describing generative AI. One of them thinks they are describing agentic AI, but they're actually just describing generative AI. They're using the term agentic. And that was like a major aha for me that, you how well have we, industry, done in even helping our customers, our colleagues, our prospects, whomever it is, the world, in really

understanding this AI thing. Like AI has existed for many many many many many years and you can it's embedded in your phone, it's finishing the word that you're typing, it's predicting a bit of it and it's existed in models forever but to your point.

this evolution of what Salesforce has gone on, we've been helping with predictive analytics with our tools for quite some time. And you're right, when we introduced Einstein within our sales platform, Einstein being able to generate summaries, that's generative AI. And being able to, I come from a banking background where the advisors, the frontline advisors, the work that they would have to do to look at 13 different screens to prepare for a call and hope that they actually do that and then hope that they've got the right information that they've

Jason Hiner (11:25.411)
Hmm.

Shibani Ahuja (she/hers) (11:39.01)
organized it and that they've extracted the insights to be able to have an assistant that's able to do that for you. That was great. Now, now the world that we're entering is a Gentic AI, which is autonomous action. It's not just ask a question, get a generated response, but with that response, is it writing directly to that CRM field? Is it creating a ticket? Is it now able to send an email to a customer that customer sees openings that this advisor is available is selecting the slot that's appropriate for them.

coming back to the advisor to say, a meeting's been booked based on availability in your calendar, all done by an agent. and I've prepared the summary notes. and here's the next best action or next best offer to talk to that customer about based on their profile. Like, that's the evolution. We've been on it fast and furious to say, we're really gunning in the agentic AI space where humans and agents are working together.

Jason Hiner (12:30.969)
Very good, okay, so it's now the agent force is the era we're in with Salesforce. So for those who are still coming to understand what you all are doing, break down the difference for me between agent force and agent force 360. These are two terms that I hear a lot from Salesforce.

Shibani Ahuja (she/hers) (12:36.493)
That's right.

Shibani Ahuja (she/hers) (12:55.039)
Agent Force, and Jason, I may need to phone a friend to make sure I've got the right explicit language for this one. Agent Force is the AI product that we have. If you're on a website and you see an agent pop up, that is powered by Agent Force, the actual product itself of an agent and the technology and the capability, and I'm oversimplifying it.

Jason Hiner (13:01.957)
Okay, okay, no worries.

Jason Hiner (13:08.932)
Okay.

Shibani Ahuja (she/hers) (13:21.679)
Agent Force 360 is an overall platform.

where agent force, to enable that one agent to do what it's doing, you actually need to have that harmonized data, that context, the syntax, sorry, the semantics, that layered on top of that governance, being able to protect all of that data, having the applications or the channels in which that agent is showing up. So it's so much more than just the creation of an agent. It's the full ecosystem around that agent to enable that agent is able to, even the flows that the agent has got the rails

Jason Hiner (13:35.588)
Yeah.

Shibani Ahuja (she/hers) (13:55.437)
When you want an agent to do more than just answer a question, so where it's accessing a knowledge article, you want it to perform an action, it needs the rails to be able to travel to perform that action. No different than a human would need to be given access rights to be able to perform XYZ operation. Agents need that full and holistic ecosystem to be able to really operate at scale alongside humans.

Jason Hiner (14:18.735)
Very good, and that makes a lot of sense. So Agent Force is almost like the brand that's describing your egenic strategy and what you do. And then Agent Force 360 is really like the suite of products, because there are a lot of things that go into creating agents. And one of the biggest challenges, and you talk about working with some of the biggest companies in the world, essentially thinking with them about agents, it starts with data, their requirements, and often very stringent regulatory requirement.

Shibani Ahuja (she/hers) (14:38.969)
Mm-hmm.

Jason Hiner (14:48.699)
requirements, governance, all of those kinds of things like the data layer, data sovereignty, all of those things. It gets super complicated. They can't just run a quick trial and move something to market really quickly because they have all of those controls and knobs they have to turn.

Shibani Ahuja (she/hers) (14:48.875)
Absolutely.

Shibani Ahuja (she/hers) (15:06.784)
But it's so funny, you're saying that, you're acknowledging it, I love that you are, yet early days when we would talk about agentic AI, and it kind of comes back to even understanding, did people really understand the definition of it, you know, the first thing I would say, I would hear is, well, I'll build my own. I'm like, what? You're gonna build your own agentic AI?

Jason Hiner (15:17.316)
Yeah.

Shibani Ahuja (she/hers) (15:23.085)
Like, what do you mean? You're gonna build your own agentic agent? How are you gonna do that? And we had this whole storyline to say an LLM is not enough because a lot of people, what they were saying, I'll build my own. In some cases, people were saying, I'm gonna build my own model. In other cases, people were saying, I'm gonna build my shell of an agent. But not again, like to your point, you're very well calling out. In order to build an eco, the environment, the holistic environment that enables an agent to operate at scale with confidence,

Jason Hiner (15:23.15)
Mmm.

Jason Hiner (15:40.291)
Mm.

Shibani Ahuja (she/hers) (15:52.947)
It's many component parts. And that's where that full and complete view of what are all of the component parts to be able to build an agent and deploy it at scale. Or you get what you kind of called out, experiments. Or you get use cases that aren't actually able to scale. That MIT study that says that 95 % of pilots fail to deliver value. And it was a big theme of I'm going to build versus buy. And I would say, OK, give it a shot.

Jason Hiner (16:15.535)
Yeah, yeah.

Shibani Ahuja (she/hers) (16:22.841)
I'll still be here when this polite polite Canadian will still be here when you fall flat on your face

Jason Hiner (16:24.282)
Yeah.

You

There's so much complexity at scale, like the complexity also scales very, very rapidly for those things. So because of that, or maybe I'm making a leap here, I think because of that, you all created this agenic maturity model to help understand this complexity and make it easier to grasp and easier to, in a sense, see where do you

Shibani Ahuja (she/hers) (16:38.787)
Yeah.

Shibani Ahuja (she/hers) (16:48.867)
Yes.

Shibani Ahuja (she/hers) (17:00.963)
Bingo.

Jason Hiner (17:01.578)
Is that right? Tell me about the Agenic Maturity Model.

Shibani Ahuja (she/hers) (17:05.197)
You've nailed it. Those are many of the primary reasons. Again, it's really like if we started talking about a product that people just didn't even understand the underlying architecture for or what it was or why it was enabled. Kind of coming back to where we started this conversation is like grounding in really the what is it? That's exactly why we created the agentic maturity model. was sometimes it's easier to help people understand what something is when you show it relative to something that they are familiar with.

Jason Hiner (17:34.252)
Mmm, you sure?

Shibani Ahuja (she/hers) (17:35.074)
did was we put it on a scale to say look AI is the scale and way on this side of that scale is chat bots that you would have had like the OG chat bots of if this then that and someone would have to configure every unique combination of what's I would be so frustrating from a customer point of view because you slightly deviate from how it's programmed to answer and it would be like does not work does not work that's like you know level zero of AI a gentic AI meaning it's not on that scale

Jason Hiner (17:54.275)
Hmm.

Shibani Ahuja (she/hers) (18:03.874)
And then you'd move to copilots. In copilots, you think about the space of when ChatGPT really first opened up, it was a great way to ask a question and get a response on a select data set, but it wasn't performing any actions. But even now, ChatGPT is able to perform those actions. So you can see that it's starting to mature into more autonomous or agentic. Or GitHub Copilot. was created, built.

for purpose in a space, in a place where it was generative. And then our agentic maturity model, those are still described as level zero, our agentic maturity model really starts at level one, two, three, four. But we created it, and level one being a simple, you ask a question, it gives you the response, but also makes a recommendation. So you can see the reasoning starting to come in. At level two, instead of making that recommendation, it's performing the action for you. Level three, it's performing action across more complex domains. And last year we used to say,

Look, level four doesn't even exist because that's where you've got agent to agent communications from different platforms. while we certainly created that to help people understand what is the, what is agentic AI on the spectrum of AI, it was certainly to help folks, help them process and digest where am I and where do I aim to be, say in a year, two years, three years. But here's a few other ways that it ended up being very valuable is like number of CIOs I would speak to and they'd be like, there's either panic or paralysis.

I've got so many use cases. I've got 170 use cases and I don't know where to start. And I would say to them, take this maturity model.

Jason Hiner (19:28.645)
Okay.

Shibani Ahuja (she/hers) (19:36.13)
start organizing your use cases by level one, two, three, four. And I'm gonna say something that is very counterintuitive to a sales organization. I will slap your hand if you try to implement a level four, three, or two use case before you've implemented a level one because start with a single use case. This is brand new technology for you. It may be within your stack. It's maybe something you're new, purchasing net new.

Jason Hiner (19:48.323)
You

Shibani Ahuja (she/hers) (20:01.518)
Take a single use case that you've aligned with the business on. Ideally, choose a use case that's going to make the lives of your colleagues better. Don't choose a use case that is the sexiest, most complex one that you want to talk about on an investor day call. Find one that's going to make your colleagues' lives easier. The frontline colleagues that have to deal with the hardest, most administrative tasks, fix that somehow.

only look at the data elements that are necessary to put that use case in production. Only cleanse those data attributes. Don't do everything under the sun. Only look at the technology that's necessary that you may have to purchase net new to get that use case in production. Then, as we've got all of this angst around we're moving to a consumption model, just put that in production and get the business who's going to benefit.

Jason Hiner (20:29.541)
Hmm.

Shibani Ahuja (she/hers) (20:46.456)
from a colleague perspective on capacity or NPS to sign up for that. Track those. Do a time study beforehand, do a time study after. Like do the basics.

Then you can start to evaluate the true ROI. You can start to see this consumption model, is it delivering value in a reasonable turnaround time? Then and when you've proven out this technology, you've really started to understand the architecture, you've understood the change management around the human impact of this. That's when you can move to level two, then level three, and then level four. So it ended up becoming so valuable from many perspectives.

Jason Hiner (21:16.153)
Yeah.

Jason Hiner (21:23.653)
that's such a good framework and such a way to break down something so complex and challenging and overwhelming to a more digestible and approachable thing. Because you cited it, that we're sort of in this shadow of this MIT survey that companies are struggling. They're investing a lot of money in AI, and they're struggling to create the ROI that they need from it and the results from it. And the pressure on that in 2026 is only going to

Shibani Ahuja (she/hers) (21:31.844)
Yeah.

Jason Hiner (21:53.728)
it up right if they've been investing money for a certain period of time. Yeah.

Shibani Ahuja (she/hers) (21:54.446)
Yes. Mm-hmm, that's exciting. And I always say like, three parts to the stool. One is share the technology. Two is how you've implemented it.

Jason Hiner (22:06.597)
Okay.

Shibani Ahuja (she/hers) (22:06.978)
And three is what are the use cases that you've sold? Have you just chosen a really bad use case where you didn't actually quantify what the value or benefit was going to be? It was just a snazzy one that you were going to be able to tell on your investor day call. Or did you get the business? I used to be in technology. Did you build a shiny object hoping the business would be signing up for the benefits, but you forgot to interlock with the business on the benefits and get them to really sign up for it? So it's looking at all three parts to make sure that you are really driving this ad

Jason Hiner (22:19.992)
You

Jason Hiner (22:31.407)
Yeah.

Shibani Ahuja (she/hers) (22:36.932)
an entire organization, as an entire C-suite collectively, while bringing your colleagues along with, not as something that you are doing to your colleagues.

Jason Hiner (22:46.969)
Yeah, so.

Agent Force is so critical to Salesforce's own strategy. You're playing a key role in this because you see and you are expressing the value, I know you can't confirm or deny this, but there was even a rumor in the last, or a report, shall we say, in Q4 that Agent Force was so important that Salesforce was considering changing the name of the company from Salesforce to Agent Force. Like I said, I know you can't comment on that, but can you talk about,

Shibani Ahuja (she/hers) (22:56.079)
Yeah.

Shibani Ahuja (she/hers) (23:08.174)
you

Jason Hiner (23:18.553)
the centrality of agent force to the larger and broader Salesforce strategy. I know other executives have talked about it, but how do you talk about it when you're talking to companies and they're trying to decide, do they want to work with you? How serious are you in this whole AI agent thing? Aren't you a CRM company? What does that look like when you communicate it?

Shibani Ahuja (she/hers) (23:39.684)
Yeah.

You know, changing the name of our products to have agent force before them, sure, that's one signal. But the real signal is you don't even have to talk to them about it. If I've got a customer in front of me or a prospect or someone is just curious to know about, like, what is this thing and is it really helpful? I pull up my own Slack instance and we launched Slackbot, which is, Slackbot is our internal agent for every colleague. Holy moly, how this Slackbot has changed.

Jason Hiner (23:59.332)
Hmm.

Shibani Ahuja (she/hers) (24:11.638)
how you do day to day. Like I'll tell you, you you kind of called it out. You're jumping from podcast, I'm jumping from podcast to like interview from customer meeting to a dinner. I have Slackbot as my immediate, I'm about to do this. What can you tell me about this customer? And how it generates a summary and then says, you want me to put it in a canvas for it to be printed out. Like that's wild. And it's accessing pools of data.

Jason Hiner (24:13.285)
Mm.

Shibani Ahuja (she/hers) (24:40.47)
that if I had to even figure out who has the information, like who can collate this for me, it would have been so painful. Slackbot has become my, I know that we've got an HR base camp where can ask questions, I just go to Slackbot. Because it's right where I work. I'm in my Slack all day long. It's a little icon. It's my cheat sheet for absolutely everything. And so we always say we drink our own champagne.

I don't think I've ever seen the company drinking its own champagne as much as it is now. Our colleague sentiment on Sockpot is there's a 96 % satisfaction rate with Sockpot. 96%. Yeah.

Jason Hiner (25:19.941)
Really, really. And for those who don't realize, Salesforce owns Slack. It's part of Slack. So that's been developed internally while it's been part of Salesforce. So that is maybe the most customer facing AI agent that you all have because Slack is used by so many businesses, especially in the US, very, very high percentage of businesses.

Shibani Ahuja (she/hers) (25:24.962)
Yes.

Shibani Ahuja (she/hers) (25:43.398)
That's right, that's right. And I could go on with examples. There's one sawdust example where I love this story. So companies like Salesforce, we put out gated assets. Gated assets could be like a PDF of information, but we dangle it to say, you only get that PDF if you give us your email address or some sort of details. It's like a lead. Well, so many people fill out those details and those are considered leads. We have AI to help us determine what are the hottest scoring leads, meaning the ones that are probably going to convert to some sort of business.

Jason Hiner (26:00.345)
Yep. Yep.

Shibani Ahuja (she/hers) (26:13.331)
We prioritize those. There's a bunch of leads that we say fall down, like they fall like sawdust on the cutting shop floor. And no, we'll never get around to calling them. So what we did internally, we said, let's experiment with something that we know is already going to the wayside. Like we're not going to be able to call these leads. We created an agent. We call it the SDR agent. I call it the sawdust agent. I love it. And this agent started to pick up these leads off the floor.

Jason Hiner (26:38.821)
you

Shibani Ahuja (she/hers) (26:42.639)
would look at the customer, send the customer an email based on the context of what they were looking at, what they were interested in, and started to actually send them emails and connecting them if they were showing interest through this email exchange and engage. Turned it into actual leads that were qualified and has actually generated revenue. This is free money, free money. This is something that we would have just let go. So the amount of marketing-driven pipe that it's generated, but the actual deals that have been closed.

Jason Hiner (27:01.989)
Wow. Yeah. Yeah.

Shibani Ahuja (she/hers) (27:11.607)
It's working. And that's again revenue generating now.

Jason Hiner (27:13.465)
Wow. Sure. That's a separate product within Agent Force? That's part of the Agent Force agent. Okay.

Shibani Ahuja (she/hers) (27:20.013)
No, that is AgentForce. That's one of our AgentForce agents. That was us drinking our own champagne and applying an SDR agent, a sales development rep agent, to pick up these leads. Ecosystem. Yeah, this is where we become customer zero is we are our own customer.

Jason Hiner (27:31.001)
Got it. Inside your own company. Very cool.

Yeah, very cool. as you're looking at and talking about, for the ones I'm sure you have examples you can talk about, what are some of the companies that you have case studies out there and examples that you can speak to that are doing some of the most interesting things with agents that you're seeing and that really kind of are at the forefront of this?

Shibani Ahuja (she/hers) (28:02.929)
At the risk of being a little bit repetitive because I'm a little excited about it and the recency of it with Slackbot.

I've talked about how we use it internally as colleagues to really enable everything we do. You may want to pay attention in the fourth quarter of Super Bowl, because you might see a tiny little thing with Mr. Beast and Salesforce about how Mr. Beast runs on Salesforce and Slack and how they're leveraging agent force. And so that's a huge story that's blowing up right now with a couple of the commercials and teasers that we put out there. But the teasers that are out there talk about how the Mr. Beast corporation

Jason Hiner (28:17.157)
Okay. Okay.

Jason Hiner (28:23.685)
Hmm.

Jason Hiner (28:34.391)
Okay.

Shibani Ahuja (she/hers) (28:39.691)
is powered by, AgentForce is powered by Slack in how they engage and communicate with one another, specifically Slack.

Jason Hiner (28:48.357)
Very good. OK, how about for 2026 as you're looking ahead? You do you think you mentioned that a lot of 2025 was explaining, you know, the the AI agent thing as you look ahead in 2026? Are there companies you don't have to say who, but really more in the broader sense and maybe in industry, industries and companies that are ready to really take this to the next level and are ready to kind of start seeing

Shibani Ahuja (she/hers) (29:04.239)
Yeah.

Jason Hiner (29:18.311)
some of that ROI. Do you see some strong signals for what's ahead?

Shibani Ahuja (she/hers) (29:24.173)
Absolutely. think that in two spaces I'm starting to see these signals. One is newer organizations that are just starting up. So I actually was fortunate enough to be in Davos for World Economic Forum and get to meet with very senior leaders of organizations of all sizes. And there was one in particular that I won't name, what they were doing is it was a co-founder that I was speaking with. And as a co-founder, you have a certain mentality to keeping costs low.

building your business, maximizing your revenue, and they hadn't yet hired all of the back end operations teams, team members that they would need as you grow. They were saying, we want to run like a lean, agentic enterprise. How do we get in on this now so that it's not that we are hiring a bunch of people and then we have to look at how we have to reassess the workforce? We want to start as a lean, agentic organization. And I think that there are many organizations right now that are more like early stages.

saying that same statement to say, do we run as a lean, agentic enterprise? Which is very much so the message that we're trying to drive is that we are, it is all about the agentic enterprise. And for larger organizations, I think...

where people were understanding what what agentic AI was, there was also a very strong association with agents, maybe AI spoken broadly, and cost reduction. Or it was all like the FTE reduction, like it was not the right sentiment of how you should be looking at it. A big shift that I've seen as of late is organizations saying, hang on a second, this can help me identify new revenue streams. This can help me if I had a contact center,

Jason Hiner (30:51.919)
Sure.

Shibani Ahuja (she/hers) (31:04.723)
that ran nine to five with this, I can run it 24-seven. And so now it's about the, if you look at the example that I shared around the sales development rep agent, hang on, this can make money off of something that we were just gonna let drop. So now organizations are starting to look at growth use cases. And in that case, they're saying, wait, this technology can help me be a bit more competitive because while I want to expand my business, I don't have to expand my resource footprint. I can augment my resources.

Jason Hiner (31:09.018)
Yeah.

Shibani Ahuja (she/hers) (31:34.643)
with this agentic AI technology that enables them and us collectively to go faster.

Jason Hiner (31:41.583)
Very cool. Since you have this background in bank and finance, there's these interesting narratives there. It's sometimes seen as something that moves very slow and is using old technology and maybe still running mainframes in some cases or that. And then at the same time, they also, in the areas of cybersecurity and other things, they're on the leading edge of many things. And so where do you see, can you speak to where is sort of financial support

Shibani Ahuja (she/hers) (31:50.417)
you

Ha ha.

Shibani Ahuja (she/hers) (31:58.373)
Yeah, yeah.

Shibani Ahuja (she/hers) (32:05.392)
Yes.

Jason Hiner (32:11.527)
services at in this whole AI revolution, aogenic AI and broadly too, because I've even heard somebody with some of the recent chat bot, not chat bot, agents, more like Clawbot and things like that, Clawbot and now Open Lobster, they're like, I don't know how some of my colleagues who are doing financial analysis are going to survive because I can give this thing, I can point this thing at a bunch of data and it just spits

out all the reports I need, you know, very rapidly. there's all these conflicting narratives about where and what's happening in financial services. We'd love your take on it since you come from that background.

Shibani Ahuja (she/hers) (32:47.951)
Yes.

Shibani Ahuja (she/hers) (32:51.931)
For sure, the juxtaposition between being risk-oriented and risk-averse and conservative versus this, well, I could give an agent to do so many things. I think that there is a desire. There's probably a financial institutions are ripe with use cases, but they're also ripe with regulatory, privacy, governance, compliance. And I think that that's where there's a bit of a healthy tension.

Jason Hiner (33:02.618)
Yeah.

Shibani Ahuja (she/hers) (33:20.077)
Now, here's a few things that I think financial institutions have done or the finance industry is before this, first of all, AI has been around for so long. If you think of financial modeling, like AI, I think predictive analytics, or even when you're looking at segmentation of customers, you know, before this whole AI hype and really when generative hit the mainstream to then agentic, before that, the hot term was personalization.

Jason Hiner (33:30.309)
Yeah. Yeah.

Shibani Ahuja (she/hers) (33:49.297)
not just for banks, but for many companies, how do we create a segment of one? How do we create omni-channel experiences for a customer in their channel of choice? Meaning, as a customer of a bank, you might choose to go into a branch, you might choose to do banking on your phone, you might choose to do it any which way by calling in. But you want a consistent experience and you want your bank to know you. That personalization architecture, and I've written about this, I have this hypothesis that...

organizations, whether it was a bank or whether it was retail or other industries, that invested in building the architecture necessary to create these personalized experiences. My hypothesis is that they have a head start to agentic architecture. So personalization architecture, I would break it down. The reason why I say this is ultimately...

Jason Hiner (34:34.597)
Okay. Yeah.

Shibani Ahuja (she/hers) (34:41.264)
to personalize to you. first, let's step back. A bank will have multiple lines of businesses. Typically, they have a commercial arm, a consumer arm, sorry, a consumer arm could be things like your checking, your credit card, your basic day-to-day banking products. Then they have typically a wealth arm, which is for like wealth advice or higher value asset management. They've got an insurance arm.

They've got maybe a commercial arm where if you're a small business owner as well, all of these arms are typically fragmented and siloed, but every arm thinks that they have the customer view. But the reality is the real customer's view when you've got harmonized data across all of these. So organizations and banks in particular that were trying to drive personalization at scale, meaning how do we harmonize all of our data to get one view of JSON across all of these lines of businesses and portfolios, they would have needed three core things to oversimplify.

They need clean, harmonized data, a single customer profile. Second thing they need, they easier said than done for sure. But if they were tackling that, check, they've got one part.

Jason Hiner (35:40.355)
Easier said than done, right? But yeah.

Jason Hiner (35:46.488)
Okay.

Shibani Ahuja (she/hers) (35:46.972)
Part two, and this is where products like Data360, doesn't matter what lake, what warehouse, what source system, which CRM, where that data is sitting, to be able to harmonize it in a way that you could extract it near real time to activate it. So I'm gonna park that. So data was one component. Second component was, now you got all this data, hopefully you've cleansed it, it's harmonized, you got a single view of Jason, you know everything about him that's necessary to know, you need a smart brain to apply what's the next best action or next best offer for Jason.

Jason Hiner (36:00.133)
Okay. Okay.

Shibani Ahuja (she/hers) (36:16.916)
that'll have the highest propensity for him to want to buy or to that based on his patterns of behavior, based on his transactions, based on his holdings of what products he holds, what is Jason going to be most interested in and it was predicting that too?

Jason Hiner (36:33.285)
Okay.

Shibani Ahuja (she/hers) (36:33.584)
The last thing, so data, that smart AI, the last thing was a channel. Okay, if we know that Jason's just gone on a mortgage calculator and straight from there went onto the mortgage product page, that should immediately fit into your data source. The brain is telling us, listen, this guy, Jason, he's interested in a mortgage. We shoot that out in every channel to say, advertise mortgage products to you because you've just given us a signal and based on your holdings, we can see that you're getting ready to buy.

to buy a house. And so that was the three things, the data, AI channels. Now if we look at what's necessary for agentic architecture or for an agent to really be able to perform actions across an organization, an agent needs data.

Jason Hiner (37:04.293)
Hmm.

Shibani Ahuja (she/hers) (37:16.964)
It needs clean data, contextual data that is going to feed the brain, which is the AI element. And the channels are where that agent is either traversing the virtual halls of an organization to perform a task, or whether it is on a website where they're showing up, that's a channel, or in my Slack as an agent, or in any internal CRM system, there's an agent. So these are the channels. So those three component parts.

If organizations like financial institutions have built this personalization architecture, it's giving them a leapfrog into agentic architecture.

Jason Hiner (37:54.905)
That's really, really great and also incredibly hopeful, I'm sure, to people that are in financial services that are thinking about this and thinking, okay, look, we've already done some of the work. We're on the path, as it were.

Shibani Ahuja (she/hers) (38:00.561)
Yes.

Shibani Ahuja (she/hers) (38:07.032)
Exactly, like, and now it's not the question of build versus buy, like don't look at these things as well, personalization is over here and our AI initiatives are over here. think organizations really need to think about how are you using AI to solve your business problems, not to just make it a headline.

Because when you're choosing a use case out of the blue for the sake of a use case to tell it at an investor day call, are you really anchoring it back to, was it personalization at scale that you were trying to solve? How can you marry the two and be able to drive your business priorities forward in a meaningful way?

Jason Hiner (38:39.791)
Very cool. So clearly you've thought about all of that at the very atomic level because of your work at Financial Services. How did you end up in Financial Services? How do you end up in tech? That's a question I love to ask people on the podcast. How did the journey wind its way to here?

Shibani Ahuja (she/hers) (38:46.116)
I have.

Shibani Ahuja (she/hers) (38:50.726)
Woof.

Shibani Ahuja (she/hers) (38:59.634)
Curiosity. I think that's it. I think the simple answer is curiosity. I again, I ended up in banking and had been in insurance and banking for almost all of my career. done a startup.

Jason Hiner (39:01.304)
Okay?

Shibani Ahuja (she/hers) (39:13.488)
But telling the customer success, like I was so anchored on the customer experience and the colleague experience, really. In any job that I've ever been in, and the last role that I was in in my bank was very much so customer and colleague focused. Suddenly Salesforce is an organization whose values I strongly and deeply believe in. Like deeply believe in. You know, they bring me in to do a customer success story. I just, thought that was, you know, when I come back to saying I gave unfiltered advice, one of the things that I said was like, you call yourselves a CRM.

company. Like you're doing yourself a disservice because to me in my organization you've become a customer engagement and colleague enablement platform. Like process that. That is so much more across sales, service and marketing. This is me as a customer. I was saying this. I was like trying to give them a wake-up call to say stop calling yourself.

Jason Hiner (39:53.349)
Mm.

Jason Hiner (39:58.905)
Yeah.

Shibani Ahuja (she/hers) (40:03.172)
a CRM company, you are so much more. And entering this space of agent AI and agent force and agents that are now truly helping colleagues and customers, I'm going, now we're talking.

Jason Hiner (40:16.101)
You've proven my point of what I said about you as a communicator. I'm sure that they heard that and they're like, we have to hire that person, whatever, I don't care what position it is, they're expressing what our company is and should be better than we are. So that's amazing. You're clearly an excellent, yeah. No, it's good. it's, you know, to be, it's really.

Shibani Ahuja (she/hers) (40:19.858)
You

Shibani Ahuja (she/hers) (40:24.306)
You

I feel very lucky.

Jason Hiner (40:38.729)
wise of a company to see that, right? To see like, here's a customer that's telling us about, you know, who we can be and who we should be in a way that's really meaningful, you know? And so kudos that they saw it.

Shibani Ahuja (she/hers) (40:50.785)
It's a special culture here. That's the other thing about the culture here is that there's no hierarchy. I can speak to anyone at any level. And it is truly encouraged. Like we are a 73,000 person startup company is the way I describe it. And it was like, it's the part that attracts me. Like one of the most attractive things about working here is the fact that we, as a 73,000 person organization, or 70,000 plus,

Jason Hiner (40:57.989)
Mmm.

Shibani Ahuja (she/hers) (41:18.156)
can operate so nimbly and adjust with where the market needs to go. It's amazing.

Jason Hiner (41:23.897)
That's great to hear. That is great to hear. Like I said, you're a great storyteller. So I want to put you on the spot and say like, what are you clearly have had an amazing journey since you've been at Salesforce, but maybe during Salesforce and even throughout your career in tech, what are some of your favorite stories in your career? That one, the one that you told is an amazing one. So because it was so good, I kind of would like to hear more. And I bet the audience would too. So what are some of your favorite stories about your journey in

in this crazy industry.

Shibani Ahuja (she/hers) (41:57.362)
I gotta say that one kind of takes the cake. Like how many times do you have Mark Benioff sitting six feet from you and 450 of his closest leadership team in the room with the entire company dialed in and someone says, hey, what advice do you have for Mark and the company? And you're thinking, yeah, it felt like comedy hours, so much fun until...

Jason Hiner (42:00.323)
You

Yeah.

Jason Hiner (42:11.363)
Yeah.

Jason Hiner (42:17.477)
and you get to unload.

Jason Hiner (42:22.147)
Yeah, amazing. How about for the past year then? Because you've gotten to have so many conversations. You're probably having three before the end of this day after we're off with other leaders. What are some of your favorite stories since you've been in the role and been doing it? What are some of the interesting things that you've gotten to do? You mentioned Davos, but I'm sure there's many other Davos's in there too.

Shibani Ahuja (she/hers) (42:42.674)
There are quite a few. are quite a few. You know, maybe I'll say thematically, like I've had the incredible fortune of traveling to Tokyo, Sydney, London, Brazil. It was, the most incredible stories are how raw, like you have these senior leaders in a room, CIOs of enterprise organizations, and how real the conversations got when you would.

when you would just break it down and kind of start with comedy. Maybe I'll give you an anecdotal example. I tease my own company about this and we're changing it. That's why I can tease. Oftentimes I get asked, will I come in to be a keynote speaker for an intimate dinner that we do? We might have about 12 CIOs. Will you come in and say a few words? And I come in and I'm quiet and I don't say much and I don't introduce who I am. There's no slide that has my name or title on it. And then as the dinner

Jason Hiner (43:13.004)
Okay.

Jason Hiner (43:19.3)
Okay.

Shibani Ahuja (she/hers) (43:38.526)
starting I go shut the door really loud and everyone's like what's going on and I go shut the door this is your timeshare pitch you're gonna listen to me talk about our products for 45 minutes and then you're gonna get a really nice meal out of it maybe you'll get a gift

Jason Hiner (43:43.685)
Ha ha ha.

Shibani Ahuja (she/hers) (43:56.242)
But that's what you're gonna get. And then I go, isn't that sometimes how it feels? It's not how it's gonna be anymore. We are a different company. We are a company that is not focused on just like making the quick sale and running out. We're a company. And so people would start laughing going, oh my gosh, you do sound like a Timeshirt. Like, you know, the past, it was a bit like a Timeshirt pitch. And now we're a company that all of our sellers are being focused on not just selling the product. Now it's on how do you sell the product and ensure customer success. But you know, it was a...

Jason Hiner (44:01.349)
Ha ha ha ha.

Jason Hiner (44:15.439)
Sure.

Shibani Ahuja (she/hers) (44:26.196)
You kind of poke fun to relate, because I thought, man, you should read some of my LinkedIn articles. I've got stories of what it felt like to be a customer. But we're changing that. We are changing all of that. But those were some of the fun moments where people were like, what is going on in here?

Jason Hiner (44:35.29)
Yeah.

Jason Hiner (44:42.309)
That's amazing. You're proving my point once again. did you, how, did you do that at a bunch of different ones? Is that like a standard playbook or you've just done that a few times? Yeah, yeah. I bet, I bet.

Shibani Ahuja (she/hers) (44:45.499)
Yeah.

Shibani Ahuja (she/hers) (44:50.032)
A little bit. Sometimes. Sometimes I do it. But the reactions are always the same.

And then they start chuckling and the head start nodding and they go, okay, she gets it. She's been on the customer side. And I go, yeah, I've been on the customer side and I promise to always meet with customers. I'm not in sales, but I promise to always stay close to the customers because again, as an organization, many of us do customer meetings so that we have our finger on the pulse. We really want to know what's working, what's not. How do we show up differently? Like this is a very different organization than the organization that was selling to me when I was on the customer side.

Jason Hiner (45:24.741)
Sure. I love that.

Shibani Ahuja (she/hers) (45:24.932)
I'm so proud to be here.

Jason Hiner (45:29.561)
What would, there's a lot in front of us in 2026 still. And like you said, it's a little bit different journey from 2025. What would success look like at the end of 2026? Like if all of these things go well, if your clients really start to implement agent force, if they start to really move their journey and agents forward, what would it look like at the end of this year?

Shibani Ahuja (she/hers) (45:57.254)
You could probably answer that from many different perspectives. I'm going to take it down to the one that matters from the heart. Right now, there's a lot of sentiment about people being worried about the impact of AI on their roles. Success to me looks like successful.

a scaled implementation of agent force and AI agents that have every colleague understanding how it's helping to up level them and how it's helping their roles and not that it is a threat. And that's going to be a long journey. It's not going to be overnight. But change is constant. that to me is what success looks like where

Jason Hiner (46:29.295)
Sure.

Shibani Ahuja (she/hers) (46:34.502)
where the tension and the worry about what this is doing to society has settled and people are starting to recognize the opportunities that this is affording absolutely everyone at every level. And that's when the change comes, when it's not just something that's talked about at the board level or at the C-suite level, but actually when every single colleague is saying, man, this, the Slackbot is amazing.

Jason Hiner (46:57.669)
I love it. Shabani, thank you so much for your time. Such a pleasure to have you. Such a great conversation. Yeah, thank you so much.

Shibani Ahuja (she/hers) (47:07.004)
Jason, it was awesome. Thank you.