The Human Side of AI: Why the Future of Sales Belongs to Those Who Know When Not to Automate

A Conversation with Gabriela M. Moreno P., AI & Automations Expert

In an era when artificial intelligence is reshaping entire industries almost overnight, it can be easy to mistake speed for progress. Products ship before they are ready. Pipelines get automated before anyone asks whether they should be. And somewhere in that rush, the human beings on the other side, customers, clients, prospects, are left wondering what happened to the experience they were promised.

Gabriela M. Moreno P. has spent the better part of a decade at the intersection of people and technology, and she has seen this story play out more times than she cares to count. She drives performance through scalable processes, cross-functional collaboration, and data-driven decision-making, a phrase that, in her hands, is not a buzzword but a genuine operating philosophy.

With over ten years leading cross-functional teams and a professional identity squarely planted at the crossroads of operational excellence and digital transformation, Gabriela brings a rare perspective to one of the most heated conversations in technology today: what does it actually mean to build AI that works for people, not just on paper?

We sat down with her to find out.

"The Same User Reactions Keep Repeating and That Tells You Everything"

The Automation Insider: Let's start with the uncomfortable truth. What are the biggest challenges the automation industry is facing right now?

Gabriela does not hesitate. She has clearly thought about this, not in the abstract, boardroom-presentation way, but in the way, someone thinks about a problem they have lived with up close, rolled up their sleeves around, and tried to fix with real teams on real timelines.

Gabriela M. Moreno P.: "To me there are two major obstacles, and they both show up constantly in real-world deployments.

The first is refinement and testing. Because the space is moving fast and competition is intense, many products are pushed to market before they are truly ready. That is why you see the same user reactions repeating, 'Who tested this?' and 'This isn't doing what it promised.' The cost of that lack of refinement is time, because teams then have to ship multiple fixes and incremental versions that could have been handled earlier with stronger validation and QA.

The second obstacle is the human factor. Even with powerful AI, people still want context, empathy, and a sense that someone understands the nuance of their situation. Some issues are perfect for automation, but others are resolved faster in a five-minute conversation with a human. And AI still struggles with exceptions and edge cases in a way that can create friction on both the customer side and the product side."

"The cost of rushing to market is not just reputation, it is time. Teams end up shipping four fixes for a problem that one round of proper QA would have caught."

There is a quiet indictment embedded in that answer, one that does not get nearly enough airtime in the industry's relentless forward march. The automation space has become so fixated on building the next capability that it has systematically underinvested in validating the current one. The result is a market littered with half-finished promises: tools that demo beautifully in a controlled environment and collapse the moment they encounter a real customer, a real edge case, or a real team with a workflow that does not fit the template

But perhaps more striking is the second part of Gabriela's answer, the human factor. It would be easy to dismiss this as sentimentality, a soft concern in a hard-nosed industry. But coming from someone who spent nearly two years engineering AI-powered tools that measurably improved operational performance, it carries real weight. She is not arguing against automation. She is arguing for knowing where it belongs and where it does not.

"Sales Teams Are Paying Attention Because the Promise Is Simple and Very Practical"

The Automation Insider: Let's shift to one of the most talked-about developments right now. How fast is agentic AI growing, and why are sales teams suddenly so focused on it?

The energy shifts slightly here. This is clearly a topic Gabriela finds genuinely exciting, not in a hype-driven way, but with the measured enthusiasm of someone who has watched a technology move from theoretical to operational and seen what it can actually do.

Gabriela M. Moreno P.: "Agentic AI has been growing for more than two years, but we started hearing about it much more loudly over the last year as more companies began including AI agents in their annual goals for different parts of the business. It moved from being a concept that lived in product roadmaps and research papers to being a line item in executive OKRs.

Sales teams are paying attention because the promise is simple and very practical: agents can navigate information across the web and different sources to build a robust customer brief or even a first-pass deck, highlighting what a customer might care about and what signals suggest opportunity. Instead of going into a pitch playing the guessing game, reps can use an assistant that saves time, surfaces insights faster, and helps them tailor the conversation, which ultimately elevates the offer and makes the sales approach feel sharper and more informed."

"Instead of going into a pitch playing the guessing game, reps can use an assistant that saves time, surfaces insights faster, and helps them tailor the conversation."

The numbers bear this out at a scale that is difficult to ignore. According to Salesforce's Agentic Enterprise Index, employee interactions with AI agents grew at an average monthly rate of 65% during the first half of 2025 alone. The broader autonomous AI agents market, valued at 7.6 billion dollars in 2025, is projected to surpass 139 billion dollars by 2033, an eighteen-fold increase in less than a decade. And a PwC survey conducted in May 2025 found that 88% of senior executives planned to increase AI-related budgets over the following twelve months.

This is not a trend. It is a structural shift. And sales, historically one of the last functions to embrace technological disruption at scale, is now finding itself at the center of it.

What makes Gabriela's framing particularly useful is the word "practical." Much of the conversation around agentic AI has stayed at the level of possibility, what agents could do, what workflows they might transform. Her framing cuts through that noise and asks the simpler, more grounded question: what problem does this solve for a sales rep standing in front of a customer tomorrow morning?

The answer, as she frames it, is preparation. Context. The ability to walk into a conversation knowing not just who you are meeting, but what they care about, what they have been struggling with, and where the real opportunity lives. That is not a futuristic vision. That is a problem sales teams have had for as long as sales has existed, and agentic AI, done well, is beginning to address it.

"The Question Is No Longer What Agents Teams Are Using. It Is What They Are Building."

The Automation Insider: That brings us naturally to the next question. What types of AI agents are sales teams actually deploying today?

Gabriela M. Moreno P.: "This question is already shifting, and I think the shift itself is the most important thing to understand. It is moving from 'what types of agents are teams using' to 'what types of agents are teams building,' because every sales organization has its own motion, its own data sources, its own workflow. And generic agents often do not fit the way teams actually operate.

In practice, you see agents supporting prospecting research, outreach personalization, scheduling and routing, call summaries and next steps, and CRM hygiene. But the common thread is that the most valuable agents are customized to the sales process rather than forced into it. Tools like Claude are making it easier for teams to assemble these capabilities quickly, so the focus is less on picking a category and more on designing the right agentic workflow that matches how the team sells."

"The most valuable agents are the ones built to fit the sales process, not the ones the sales process had to be reshaped to accommodate."

This distinction, between using and building, is one of the more nuanced observations in the current AI conversation, and it points to a maturation happening in real time across the enterprise. The first wave of AI adoption was largely plug-and-play: organizations grabbed a tool off the shelf, ran it against their existing process, and measured what happened. The results were mixed, precisely because the tools were designed for a generic version of a workflow that rarely matched any specific organization's reality.

What Gabriela is describing is a second wave, one where the teams with the most sophisticated AI practices are not just consumers of agentic capabilities, but architects of them. They are mapping their own sales motion, identifying the friction points, and then designing agents that address those specific points, rather than retrofitting their process around an agent's limitations.

This has significant implications for how organizations think about AI investment. The question is not just which tool do we buy, but what do we want to build, and do we have the internal capacity to build it well? For sales teams without that capacity, the risk is real: adopting a generic agent that creates as much friction as it removes, and then drawing the wrong conclusion about whether AI works at all.

"Human Time Should Be Reserved for Where It Actually Moves the Needle"

The Automation Insider: Looking ahead, which parts of the sales process do you expect AI agents to fully take over in the near future?

Gabriela M. Moreno P.: "In the near future, I expect AI to take over, or at least heavily automate, the most repeatable and time-consuming parts of sales. Especially the work that is currently manual, templated, and easy to standardize.

That includes building first-pass sales decks and account briefs, supporting prospecting and list refinement, automating follow-ups, and assisting early-stage qualification by helping identify intent and filtering out low-quality leads. A lot of tools already do a strong job generating materials quickly and summarizing information at scale, so the natural next step is for agents to orchestrate those tasks end-to-end.

And when that happens, when those tasks are handled, sales teams get to focus their human time where it actually matters most: complex discovery, relationship building, negotiation dynamics, and any situation where nuance, trust, and exception handling decide the outcome. Those are the moments where a human in the room is not just nice to have. It is the whole game."

"AI should handle the work that is manual, templated, and easy to standardize. That frees human time for the moments where nuance, trust, and exception handling decide the outcome."

This is perhaps the clearest articulation of the human and AI division of labor that the sales industry has been circling around for years, and it is a more useful framing than most. The debate has too often been structured as a zero-sum question: will AI replace salespeople? Gabriela's answer reframes it entirely. The question is not replacement. It is reallocation. It is deciding, with intention and clarity, where human attention creates the most value and then protecting that space fiercely while letting automation handle everything else.

The data supports this orientation. Sales organizations using AI agents report productivity increases of 25 to 47% from time savings on repetitive tasks alone, according to Vellum AI's 2025 analysis. More than 80% of sales teams using AI report increased revenue, compared to 66% of those operating without it, according to Cirrus Insight. These are not marginal gains. They represent a meaningful competitive advantage for organizations that get the allocation right.

But getting it right requires exactly the kind of judgment Gabriela has spent her career developing: the ability to look at a process, understand where human presence adds genuine value, and resist the temptation to automate simply because automation is available.

Key Takeaways

What every operator, sales leader, and AI practitioner should carry away from this conversation:

Speed without validation is not a competitive advantage. Rushing products to market before they are ready does not just damage user trust, it creates downstream costs in patches, fixes, and lost credibility that dwarf whatever head start was gained. The organizations winning in automation are the ones investing in QA as seriously as they invest in shipping

AI and human support are not competing, they are complementary. Some problems belong to automation. Others belong to a five-minute phone call. The skill is knowing which is which and building systems that route intelligently between the two rather than defaulting to one mode for everything.

The shift from using agents to building them is the defining signal of mature AI adoption. Generic tools produce generic results. The sales teams extracting the most value from agentic AI are designing workflows that fit how they actually sell, not reshaping their process to fit the tool.

AI's highest-value role in sales is liberating human attention, not replacing it. Prospecting, account briefs, CRM hygiene, follow-up sequences, these are where AI earns its keep. Complex discovery, trust-building, negotiation, and exception handling, these are where humans earn theirs.

The organizations building now will be the ones leading in 2027. With the agentic AI market on track to grow eighteen-fold by 2033, the window for thoughtful, strategic adoption is open but it will not stay open indefinitely. The time to design the right workflow is before the pressure forces a hasty one.

About the Expert

Gabriela M. Moreno P. is a manager and digital transformation leader with over ten years of experience leading cross-functional teams at the intersection of operational excellence and artificial intelligence. .

Her professional focus spans AI strategy, operational process design, data-driven performance management, and the leadership dynamics required to bring cross-functional teams through technological change without losing the human thread that makes those changes stick

Disclaimer:

The views and opinions expressed in this article are solely those of the author and do not reflect the official views, positions, or policies of his employer or any affiliated organization. The content is provided in a personal capacity and is not intended to imply any form of official representation or endorsement.