AI That Works: Enterprise Adoption Insights From the Field

This is a summary of an episode of Pioneers, an educational podcast on AI led by our founder. Join 3,000+ business leaders and AI enthusiasts and be the first to know when new episodes go live. Subscribe to our newsletter here.

TL;DR:

• Enterprises are moving past AI pilot purgatory into real deployment, driven by frustration with slow-moving incumbent vendors.
• Underwriting is emerging as a top use case, where AI agents can preserve institutional knowledge and boost productivity.
• Agentic AI is not just automation—it enables orchestration, augmentation, and better integration with existing systems.
• Safety, compliance, and integration are top priorities when enterprises evaluate AI vendors.
• Startups with domain expertise and forward-deployment models are gaining an edge over traditional RPA vendors.

Before we dive into the key takeaways from this episode, be sure to catch the full episode here:

Meet Suzanne - Chief of Field at Multimodal

Suzanne Rabicoff, Chief of Field at Multimodal, works with enterprise insurance and finance leaders to help them navigate AI adoption beyond the hype.

She began her career as an enterprise analyst, transitioned into customer success in frontier tech startups, and later led business operations for scale-ups. Today, she partners closely with technical founders to translate AI innovation into real enterprise impact.

Suzanne coined the role “Chief of Field” to bridge the gap between visionary products and enterprise adoption, ensuring AI solutions align with operational pain points, not just demos.

At Multimodal, she champions the adoption of agentic AI that enables enterprises to automate, orchestrate, and elevate their workforce—freeing people from repetitive tasks and embedding institutional knowledge in digital agents.

From Pilot Purgatory to Production: How Enterprises Are Scaling AI

Suzanne Rabicoff describes a clear shift happening in enterprise AI: companies are tired of endless pilots that fail to scale.

She refers to this phase as “pilot purgatory,” where projects stall due to complexity, compliance concerns, or lack of clarity on ROI. Now, however, there’s growing urgency and internal pressure to move beyond experiments into real-world deployment.

‍AI vendors who once led with “visionary demos” are being replaced by teams that bring domain expertise and understand how to deliver repeatable, production-ready solutions.

Enterprises are asking harder questions—about integrations, safety, and execution.

The winners in this shift are startups that can support AI agent rollouts quickly and prove operational value across underwriting, claims, and customer experience, not just prototype potential.

Why Underwriting Is Ripe for Agentic AI Transformation

Underwriting is emerging as one of the most compelling enterprise use cases for AI agents, according to Suzanne.

‍In today’s market, underwriters are expected to do more with fewer resources—reviewing more submissions, navigating fragmented data, and meeting productivity goals without sacrificing judgment.

Suzanne notes that agentic AI is uniquely positioned to handle the repetitive, documentation-heavy parts of underwriting—things like gathering insights, structuring documents, or following up.

AI agents can also act as “intelligent force multipliers,” allowing teams to retain speed and quality without adding headcount.

‍“The underwriter is trying to do more with less. The agent becomes a digital teammate.” — Suzanne Rabicoff

Importantly, this approach helps capture and preserve institutional knowledge, embedding it into agent workflows and reducing onboarding time.

Suzanne calls this the transition from “junior” agents to underwriting support agents that become integral parts of the team.

Inside the Shift: From Chatbots to Agentic Process Automation

‍Suzanne distinguishes between traditional chatbot-based automation and agentic AI. While many enterprises tried chatbots or RPA tools to automate simple tasks, the complexity of enterprise workflows demands more flexible, intelligent systems.

AI agents are emerging as an evolution: they not only respond to inputs, but also take initiative—reading unstructured documents, initiating actions across systems, and coordinating follow-ups.

Suzanne calls this “agentic process automation”—a move from basic task automation to orchestration.

“The agent can structure, draft, and follow up. It’s not just task completion.” — Suzanne Rabicoff

‍This shift enables businesses to go beyond static flows and into AI systems that understand business logic, context, and execution paths.

‍It’s a foundational change in how work gets done—replacing the idea of automation as a “plug-in” with automation as a dynamic participant in operational strategy.

Safety First: What Enterprises Want From AI Vendors

As enterprises scale AI, safety, compliance, and governance are at the forefront. Suzanne emphasizes that enterprise buyers no longer just ask for AI capabilities—they ask how those capabilities are controlled.

Companies want clear answers to questions like: Who does the agent talk to? Where does it pull data from? What systems does it have access to?

AI vendors that come in with flashy demos but no coherent answer on safety get filtered out quickly.

Suzanne notes that successful vendors must now lead with enterprise maturity—offering controls, logging, human-in-the-loop structures, and clear delineation of what the agent will and won’t do.

Enterprises are not experimenting for fun—they’re deploying with caution. Trust and transparency are table stakes in the agent era.

RPA Isn’t Enough Anymore, And Enterprises Know It

According to Suzanne, many enterprise leaders have realized that RPA (Robotic Process Automation) alone can’t meet today’s operational demands. RPA tools were designed for predictable, rule-based workflows, but real-world enterprise processes are messy, dynamic, and data-rich.

Suzanne notes that insurance and finance teams are turning away from monolithic RPA vendors in favor of AI agents that can reason, adapt, and handle unstructured information.

These agents don’t just automate clicks—they understand documents, make decisions, and work across systems. Suzanne calls this the shift from “task execution to orchestration.”

“RPA was good for clicking buttons. But it couldn’t think.” — Ankur Patel

The future isn’t scripted bots—it’s autonomous agents that operate intelligently in complex, real-world workflows. Enterprises that want to stay competitive are actively replacing legacy RPA with AI that can learn and scale.

The Urgency to Act: Why Enterprises Can’t Afford to Wait

Suzanne describes a growing impatience inside enterprise teams. Leaders are no longer content with “research mode” when it comes to AI—they want deployment, outcomes, and momentum.

She notes that there’s real urgency now driven by internal pressure, market competition, and economic constraints. Enterprises are realizing that waiting another 12 months could mean falling behind.

“You don’t want to be twelve months behind just because you’re still exploring.” — Ankur Patel

Teams are tasked with “doing more with less,” and agentic AI offers a way to scale operations without scaling headcount.

Suzanne also points out that many leaders have already seen what AI can do, but they’re stuck trying to operationalize it.

The enterprises moving fastest are those that commit to cross-functional collaboration, work through compliance early, and choose vendors that understand both speed and safety.

How AI Agents Help Capture and Scale Institutional Knowledge

One of the most powerful but often overlooked benefits of agentic AI is its ability to retain and scale institutional knowledge.

Suzanne explains that when employees leave or change roles, their workflows, tips, and context often vanish with them.

AI agents that are trained on internal processes and documents can preserve this institutional memory, ensuring that “how we do things” doesn’t get lost. Suzanne likens this to building digital teammates who can be trained once and reused across the org.

This isn’t just a productivity win—it’s a continuity strategy. Enterprises no longer have to start from scratch when someone leaves.

Instead, they can embed expertise into the systems themselves, making onboarding faster and execution more consistent across departments.

What Visionary Leaders Understand About Agentic AI (and Why It Matters)

Suzanne points out that the most visionary leaders don’t view agentic AI as a tech trend—they see it as a new operating model.

These leaders are thinking beyond task automation. They ask: What happens when we deploy AI agents that don’t just assist people, but operate alongside them? Suzanne explains that these leaders understand that AI is not about replacing people but augmenting them with workflows that are faster, smarter, and more resilient.

They think structurally: how do we redesign teams, reallocate resources, and rethink value delivery?

This shift isn’t about “adding AI,” but rather about adopting a new lens for productivity and scale. Suzanne calls this a leadership trait: the ability to move from experimentation to transformation with clarity and urgency.

Would you like to learn more about enterprise AI implementation? Check out this episode on AI implementation and how to avoid AI deployment pitfalls with Bjorn Austraat.