Artificial intelligence (AI) is no longer a futuristic concept but fast-becoming an integral part of how businesses operate today. But there’s a new evolution emerging that’s set to transform everything from how we interact with data to how decisions get made: Agentic AI. Luca Benini was invited by Navirum to join their CEO Rory Galvin on their Exec Talk podcast to to explore how agentic AI is reshaping financial services, particularly within the Salesforce ecosystem.

What follows is a summary of the discussion. You can view the full podcast on YouTube or listen on Spotify.

From Nativevideo to Intelligent Agents

Rory asked how Nativevideo evolved into offering Vinton. Our journey started in video and audio solutions on Salesforce, initially focused on embedding video capabilities directly into Salesforce, and the platform evolved in tandem with AI advancements especially with the emergence of large language models like ChatGPT.

“We saw a natural opportunity,” Luca explains. “Our platform already transformed video and audio into text. When GPT became publicly available, it aligned perfectly with what we were building.” What emerged was Vinton: the first AI note-taker built natively in Salesforce, designed not just to capture conversation data but to act on it intelligently.

What Is Agentic AI?

Traditional software operates in a predictable, input-output model. You provide data, follow structured steps, and get a result. Agentic AI flips this model. It interprets loosely defined instructions and takes autonomous actions to achieve outcomes. “Think of a self-driving car,” Luca says. “You can’t pre-program for every scenario. The car must interpret what it sees and respond appropriately.”

In enterprise systems like Salesforce, agentic AI acts like a proactive assistant. Instead of manually combing through dashboards, users can simply ask, “What’s the current status of our projects?” The AI agent not only retrieves the answer but interprets the data, compiles a summary, and recommends actions.

Real-World Application: Weekly Digests for Executives

One of the most practical implementations Luca described was a use case in a large manufacturing firm. Executives needed a weekly digest that included updates across all open projects and service tickets. Historically, this meant sifting through spreadsheets and status reports. Now, Vinton’s AI agent compiles this data automatically, pulling from recorded calls, CRM entries, and even video meetings, to generate an executive summary complete with action items.

The result? Executives receive a clear, concise overview without lifting a finger, allowing them to focus on decision-making rather than data collection.

Centralising Data: The Unsung Hero of AI Success

The secret to unlocking AI’s full potential lies in data centralisation. As Luca and Rory discussed, many companies have valuable data scattered across tools like Jira, accounting platforms, and communication channels. Without a unified data layer, AI can’t connect the dots.

Salesforce’s Data Cloud is emerging as a critical piece in solving this challenge. Instead of asking businesses to abandon their favourite tools, Salesforce now focuses on integrating and exposing that external data to agentic AI. This approach allows AI agents to access and interpret multiple data sources without forcing companies into a single software stack.

Three Pillars of AI Value

Luca shared an anecdote from sales training he had years ago, that there are only three reasons why any business buys technology:

  1. To save money (efficiency)
  2. To make money (effectiveness)
  3. To reduce risk (de-risk operations)

These benefits are especially critical in regulated industries like financial services. With so much data to process and so many compliance boxes to tick, intelligent automation isn’t a nice-to-have, it’s a strategic imperative.

Use of Agents in Financial Services could save time by automating routine tasks like updating records and summarising meetings. They might boost effectiveness by enabling better, faster decision-making through intelligent insights. Or perhaps they help to reduce risk by flagging inconsistencies or compliance issues before they escalate.

Starting the AI journey

For businesses wondering where to begin, Luca offers simple advice: start by asking questions. “People are already using ChatGPT in their daily lives,” he notes. “Bring that same mindset to your CRM.” Start by asking your systems basic questions. If they can’t answer, it’s time to think about integrating tools like Vinton or Salesforce's Agentforce to bridge the gap.

As companies progress, they often realize they need more data sources to get better answers. That’s when the journey becomes truly agentic. AI doesn’t just answer questions, it begins to anticipate them, acting like a virtual assistant for every part of your business.

Final Thoughts: Vision Meets Practicality

Salesforce continues to lead with bold visions and futuristic demos, and it’s up to partners like Vinton and Navirum to make those visions real for everyday clients. As Luca puts it, “It’s our job to take that vision and translate it into solutions that solve real-world problems.”

Agentic AI isn’t just a buzzword. It’s the foundation of a smarter, faster, more intuitive business environment. And in industries like financial services, where the stakes are high and the data is vast, the timing couldn’t be better.

View the full podcast on YouTube or listen on Spotify.

You might also be interested in