Decision Intelligence for Revenue Leaders

Julia Nimchinski:

And welcome back to the show, Tuba Garaz, founder and CEO of Amoeba AI. You have a massive product release, correct me if I’m wrong.

Tooba Durraze:

We time our product releases, yeah, essentially. Yeah, we have, well, obviously in this day and age, there are a ton of product releases happening all the time.

I think one of the things for us in, like, this V2 of Amoeba as a company is… realizing the importance of strategic decision-making and how would AI support, not just in helping you make decisions, but making you a better decision maker in general. So that’s kind of what, we’re excited to talk about today.

Julia Nimchinski:

Let’s do it.

Tooba Durraze:

Okay, well, thanks again for having me. I think this is, like, the third time today, so folks have seen a lot of me today, but just to give people context into myself, so my name’s Tuba. The race, I am a… the founder and CEO of Amoeba AI, which is a neurosymbolic decision intelligence platform.

All the jargon words combined, which I’ll dive into in a moment. I am a scientist, an AI data scientist by background, so I tend to lean a little bit more, into architecture, infrastructure, etc.

But one of the things that us as a company have been obsessed with is how do we take all the information that’s being created and make it easy for leaders to, make their decisions.

And the big theme to that is cognitively, you’re doing a lot all the time, so how are you able to reduce your cognitive load without missing the things that are, important or needing to be made a decision about? So, we currently focus on go-to-market as a space, so generally you would think if you hold a number.

in the go-to-market space, we are the right tool for you. I’m going to share my screen, and… The academic in me, it behooves me to not talk a little before I jump into the demo. So, again.

We’re the first decision intelligence platform, so, you know, not a dashboard, copilot, chat with your data, all of those things are really useful, but basically, our world is a system that’s designed to surface what matters before you or your leadership team asks. So, how do we get away from this idea of, like.

the only way I extract value out of my information is by asking it a question. And the big tether here is, that you want a system that’s able to look over your entire business. Right?

And we’re not talking about just, like, business data, like, in the traditional sense, but the entirety of your business, understanding your goals, the objectives that you’re trying to reach, and then does all the work in the messy middle, essentially, to help you make the decision to drive those outcomes.

So, the reason I bring that up is because, When you ask. a data question, as an example, that’s untethered from an objective.

You can get all sorts of, like, interesting musings about, this could be interesting, or that could be interesting, that exists in your information, but For us, our focus is tethered to, objectively goals, and, there’s one, like, alpha goal, right, for everyone, which is, what do I do to make my business grow? So, like, even at that level.

Practically, what does that mean? We kind of, if you think of… the world in these three layers, there’s your data layer, and now the data layer obviously has your, structured and unstructured data, right? So any kind of operational system that’s, like, generating data, your CRMs, your ads, your gongs, like, all of that.

And then there is a measurement layer on top, essentially, which is basically to find out, what is happening. So A plus B equals C is where I would put that. Metrics supporting your AI-driven workflows, they kind of sit in that world.

So anything from BI to reporting to kind of things that you do with, Claude or co-worker, kind of sit in, like, those two buckets, right? Essentially, you’ll find something in your data, you’ll figure out what to do against it based on what the numbers are saying, you’ll dispatch your agents to go and do those things.

In the perfect world, it reaches a threshold, and you’ll ask them to go and do those things instead of, like, having to actually set each of these things up manually. A lot of attention went between those two layers, because folks were like, oh, without the right data, obviously our agents can’t work well.

Or folks said, if I don’t have the right context, which is a better understanding of our business, then, you know, how am I supposed to deploy my agents? We worked on the further up top, the tippity top of the pyramid, which is the decision there. So based on everything that is happening, against the objectives that I’m trying to reach.

what are the decisions that I need to make today? to impact my goals. So, not, you know, next quarter, we potentially can do XYZ to meet our goal, but, like, today. What do I do today to meet that goal?

And if you think about decision architectures, and I’m going to flip to product, but if you think about decision architectures, essentially, You can… first… in a normal organization, you can bucket them into decisions being made at the level of, you know, these different, like, areas of focus, so… you know, their marketing teams, their revenue teams, their customer success teams, their product teams, etc.

But this is a representative example of, like, how most organizations operate. So decisions, one, happen kind of deep in those verticals. And then there are decisions that happen horizontally. If I take a typical marketing org, the first thing I’d want your attention to is, think of the levels, right?

So, in any org, this marketing org being an example of this. There is your operator level, so again, an operator here could mean a bunch of people who own a particular objective or an initiative, right? In our case, that’s this level. There is, like, the middle level, so folks who own teams of people who are owning objectives.

And then there is, like, the higher level, which is the executive level. The need for What is important changes at all of these levels, because the jobs to be done at all of these levels are very different.

So, if you think of the operator level, when it comes to data-driven decision-making, the job to be done is, hey, tell me, like, I need to know what’s important to do something against that now. if you think a level up from that, the job changes a little bit.

I need to understand how my teams are operating, whether the things that are happening are actually moving goals forward, and, you know, where do I see gaps that I need to ask my team to kind of fill, right? So it’s more of like a, you are a… orchestrator, and also, like, an assigner.

So if I think of, like, composers and orchestras and stuff like that, that’s basically what you’re doing. And then, a level above that, so if you think of, kind of, your C-suite, your VPs, etc. the role is even more different. Your role is much more about observability, so you need to understand what’s happening.

But, you also want to understand where are the right points of intervention for you, because, again, your time gets limited the higher up you go. So, it’s no longer a question of, like, are my teams… obviously. are we meeting our objectives?

This is my corollary to our team’s performing, but it’s also about where does intervention for me actually move the needle? Could be on a very micro level of, like.

intervention in one particular deal, like, as an executive sponsor, could be on a more thematic level, like we’re seeing something emerge out of this, like, competitive lens that we need to get ahead of.

And then could also be on a, like, a much more strategic level, which is, hey, based on everything that we’re seeing, we need to plan for, basically, a brand redo in the following quarter. So that’s kind of what Amoeba has now, our decision intelligence, has architected around. So if I, quickly go into a couple of these, so again, you know.

we’re, like, cutting noise the higher up you get, but at this level, where I am a, I’m an operator, and I’m running, kind of, budgets… I’m looking at budgets of budget reallocation, my world looks like, okay, you know. if display is the main source of conversion erosion, hey Amoeba, watch it for me.

So, and I’m using Amoeba as your guide, or the agents in Amoeba, to basically further affirm or change the perspective on what’s happening. if I am… a… like, a team level here, then I am getting an idea of, like, how are all of my different initiatives performing?

And again, I say the word initiative because I think oftentimes when we think of measurement, you get stuck in the quantitative, but the way organizations operate is a good mix of quantitative and qualitative combined, right?

Obviously, there’s a metric to move forward, but there are qualitative things that are happening, or things that can happen, that contribute towards that thing moving forward, which is what attribution’s been trying to solve all this time.

So again, at this level, I’m looking at what is the business impact that my, my particular initiatives are happening… are having. I’m interested in what decisions I need to take. I’m also interested in what decisions does my team need to take, and interested in assigning them, right?

I may be interested in looking overall at this level, okay, what are kind of the risks and opportunities that are essentially emerging? And then, you know, I may or may not be interested in a couple of questions here and there, which is kind of like our perspectives thing.

And then at the highest level, again, it’s, like, a little bit simplified, it’s more about, kind of, observability, so maybe metrics, etc, don’t change if I’m moving up in this nesting, but more and more at this level, for a CMO, it should be as simple as reading one brief, this is a part of, like, our intelligent brief, and being perfectly aware or in tune with what’s happening.

Now, because we are, obviously, our foundation is Agentech, there are a bunch of things that you can do with these things, right? So, as an example. If I am a CMO, I’m not in the platform consuming this, I’m getting this in my Slack, I’m getting this in the morning, I’m actually not even reading it, I’m listening to it.

A lot of Farsi-sweet just, like, listen to it. But I may, from Slack directly, or from my email, then dispatch Amoeba to then go and investigate on something that I found was interesting. If I am at this level, I’m going to look at this. And I will bring it up in the next kind of team meeting that is happening, right?

So again, if you think of all the spaces that decisions are made, they kind of get collated in here, including kind of sessions, like, I’m having meetings around this topic, I am running reports or, like, deep dives on this topic. I am investigating in data, what does this topic mean? I’m looking at visual trends to see, could I pick up something?

These are all areas where decisions are made, and they kind of compose into this one hierarchy. The one thing I will say, one realization we have had is, I know the concept of a, An org chart, which roughly this represents, is… seems very, kind of. antiquated when we’re in the world of graph representation, right?

So if you think of, like, Obsidian, if you think of nodes, etc. But… The problem with, sometimes with nodes, and, like, this is something where you can probably think of it as, like, a node-based architecture, or, like, a hierarchy-based architecture in this way.

The problem with notes sometimes is it flattens the hierarchy, and which works in some cases, and it doesn’t work in other cases as well. So, this is how organizations are still kind of structured or anchored. So this is how your decisions are made.

Hence, this is how the information needs to flow up or down, in order to get you to your best results. Let me pause there. I don’t know how we’re doing on time. I think we’re kind of off time now.

Julia Nimchinski:

This is fascinating. We’re gonna prolong your session and prolong the next one, so… all good.

Tooba Durraze:

Okay, so, again, like, I’ll wrap up with this, saying that, you know, marketing’s, again, one world, revenue is one world, customer success, product, etc, etc, is one world, but in any one of these orgs, like, you can imagine there are multiple trees, so just because this tree is representative of, like, the average kind of marketing org hierarchy doesn’t mean that your structure has to be nested like this.

Like, an interesting way to nest a structure that we’ve seen folks do is making it around, strategic initiatives. Strategic initiatives are, like, cross-cutting initiatives that kind of run for, like, an entire half of the year, as an example, and not particularly a quarter.

Or structuring it around things that are hard to measure, very hard to measure, like share of voice. competitive sector, right? So what’s happening in the market that is… driving some sort of a result when it comes to, my objectives, whether it’s, like, revenue or pipeline.

But the biggest thing here is, like, even though you are looking at it at this level, none of the decisions across all of these that are recommended, any insights or recommendations. are at, like, opposed to, the directors and initiatives that might be in the sales team.

So, the hidden magic behind the scenes is at the end of the day, like, this will solve the problem of you’ll never be in a world where, you know, you, say, are meeting your pipeline number, and it’s great, but the revenue number isn’t being met. It’s meant to play all of these different decision trees to kind of flow together.

So, the perfect go-to-market orgs, the perfect, like, leadership org, essentially, gets the best of cognitive function via the compute. And the operators and the leaders are also using Amoeba as a… as an aid to be better about your decision making.

A really neat way that folks use this right now is, instead of coming in every day and kind of picking, you know, like, hey, I want to do something with this, or, like, life cycles for insights and recommendations.

is they… they have a sprint with Amoeba every 7 days, so Amoeba schedules a sprint with them, and then in Slack, or sometimes in sessions, Amoeba will just, like, think of your data scientist talking to you for 10 minutes. I’m seeing this, I’m seeing this, you’re conversing with it back and forth.

You can use voice, you can use… obviously, you can chat to it as well. You can do it in Slack, you can do it in the platform, you could do it in Teams, you can even send an email back to Amoeba to say, hey, send me an update. And then, this is what I thought about the update.

And then this becomes your personal data scientist in this, and all the way up. I’ll pause here.

Julia Nimchinski:

Love the tech, Tuba. I want to make sure that we address a couple of questions from the audience, if you don’t mind. The first one, how much historical data do you need before the recommendations become reliable?

Tooba Durraze:

It is… false to say recommendations are almost always entirely dependent on historical data, especially in the world that we’re living in right now, where everything is changing so fast, right? As an example, like, market behaviors, etc. So, if you have a ton of data.

Like, in some cases, like, folks have, all the way back, 5 years back in some cases. That’s great. If you don’t have a ton of data, it’s fine. It can work through it as well. the system does a… the caliber of insights stems out of giving the system purpose to be like, this is the objective we’re trying to reach, right? Essentially.

And then you working with the system, again, if you have context graphs, great, it plugs into this. If you have ontology, semantics. Great, it plugs into this, but the idea is, like, if you… even if you have nothing, pointing the system at the right direction is what causes everything else to flow.

Julia Nimchinski:

One more question. If Amoeba tells me to shift budget or pipeline strategy, how can I validate that recommendation before acting?

Tooba Durraze:

Yeah, so directly in Amoeba, because what happens is, like, if a data scientist says this to you, the normal business leader says, well, can you run a simulation or an experiment for me to understand? So directly inside Amoeba, if it gives a recommendation, let’s say you’re unsure about that, because this is, like, your day 5 or 6 of trying it.

run a simulation, and let it give you… when we run simulations within the platform, it gives you scenarios. You can implement it this way, A, B, or C. It is always meant to, help you within certain boundaries to understand what are the best things that you can do.

But that doesn’t mean that you shouldn’t poke holes at it with it, for it to tell you why it’s basing its assumptions that way.

Julia Nimchinski:

Love it. Thank you so much. And what’s the best next step? To reach out to LinkedIn? Go to Amiba AI?

Tooba Durraze:

Yeah, reach out to me on LinkedIn. I like geeking out these days over, like, workforce innovation organizational structures. And yeah, follow us on our LinkedIn page, or you can find more stuff on our website, amoeb.ai. Thanks for having me.

Julia Nimchinski:

Fantastic. Thank you, Tilva.

Tooba Durraze:

Bye.