How Acuity is Rewriting the Rules of Financial Intelligence with Agent Fleet

Jon O’Donnell is the Chief Operating Officer at Acuity. As part of Acuity’s Executive Committee, he works with senior leaders across strategic initiatives. Jon also oversees Acuity’s strategies, technology and digital solutions and legal aspects.

A proven business leader, Jon is passionate about applying technology to business practices and problems. Before joining Acuity, he worked with IBM, where he successfully led the Business & Technology Consulting and Cloud Consulting divisions (15,000 practitioners and $6bn annual turnover), delivering impressive market growth. For five years, he was also a member of IBM Consulting’s Global Leadership Team and held critical leadership positions in numerous markets, including the UK, EMEA and North America.

In this Q&A with Cybermedia publications - Dataquest and CiOL- Jon O’Donnell, shares more insights about the company's new agentic AI platform - Agent Fleet, which he believes could fundamentally reshape how investment banks, asset managers, and private equity firms operate. Excerpts.

Acuity has long been recognized for its deep domain expertise in financial services. What inspired your move into building a proprietary agentic AI platform, which is typically seen as the domain of hard-core tech firms?

Our core mission has always been to deliver excellent solutions and services to our clients in the financial sector. Traditionally, this has been heavily people-led, with support from technology. But as AI and automation have matured, we saw a clear opportunity to enhance our delivery model by putting technology at the heart of what we do.

That’s how we arrived at Agent Fleet, our proprietary agentic AI framework. We aren’t moving away from our domain-led model; we’re enhancing it. Our more than 6000 domain experts in banking, asset management, and private markets now use these advanced AI tools to increase speed, quality, and depth.

The key inspiration was to amplify the human expertise we already had. With Agent Fleet, we can do routine tasks more efficiently, freeing our analysts to focus on higher-value work such as deeper assessments, insight generation, and strategic support—work they enjoy more and that drives greater value for clients.

How does Agent Fleet align with your broader business goals and impact your engagement model with clients?

The core of our business, delivering deep domain expertise, hasn’t changed. But now, by layering AI into our workflows, we can offer clients a broader set of engagement models. They can continue with traditional analyst support, but they can also opt for outcome-based models, such as paying per company profile, credit risk report, or financial benchmark.

We’re also seeing the emergence of a parallel revenue stream from technology, access to our AI tools themselves. Clients can interact directly with our platforms to generate outputs or integrate the tools into their own systems. This flexibility is a major differentiator.

More importantly, our hybrid model, a combination of human expertise and AI, is what sets us apart from pure tech players. Our tools are trained by experts who live and breathe financial analysis. That combination is hard to replicate.

How do your AI agents differ from general-purpose large language models like ChatGPT? What specific value do they bring to financial organizations?

General-purpose generative AI tools are prompt-driven. You give them a query and get a response. They’re great for speed and can improve efficiency by 15–30% in many areas. But they lack rigor, consistency, and often hallucinate. They also don’t complete entire tasks or adhere to audit or compliance standards, these are critical factors in financial services.

Our Agent Fleet is built on Agentic AI principles. Instead of just responding to a prompt, these agents are task-driven, they break down a task into subtasks, assign them to specialized agents, and then reassemble the verified output.

Just to give more context, each agent has a "persona", it might be a data analyst, portfolio manager, or verification auditor. They are trained on structured data, unstructured content like PDFs or web pages, and client-owned proprietary data. Critically, we deploy 3–5x more quality assurance agents than content generators. That’s how we ensure accuracy and eliminate hallucinations.

This is especially vital in banking, where even minor inaccuracies can lead to regulatory or financial repercussions. Our approach brings rigor, repeatability, and reliability to workflows like company profiling, risk assessment, and due diligence.

What if your agents need to access gated or confidential financial data during due diligence? How does that work, especially with client-specific restrictions?

All our analysts work within the client’s environment, with their data. We don’t bypass or externally scrape gated data unless the client grants explicit access. Our tools are deployed on client-authorized infrastructure, and the agents point to the data sources clients already use.

We simply bring automation, speed, and intelligence on top of that data. We’re not introducing new data privacy risks, we’re respecting and extending the client’s existing security and governance protocols.

Do you have any early client success stories or metrics that validate the effectiveness of Agent Fleet?

Yes, many of our clients are already using our AI tools. Even before the formal Agent Fleet launch, we had AI components embedded in client workflows. Today, nearly 50% of our analysts and clients use these tools.

Clients have seen measurable gains in productivity. Tasks that once took 60–90 minutes can now be done in 10–15. But the more meaningful impact is qualitative, it frees analysts to think deeper, to generate insights, to explore scenarios, and support decision-making in a richer way.

We haven’t seen clients asking for fewer people. On the contrary, they ask us to do more—go deeper into due diligence, explore new asset classes, or build richer reports. That’s the real “proof of the pudding.”

There’s a larger narrative around AI causing job displacement. From your perspective, how do you manage the human-AI balance in your workforce?

That’s a fair concern and one we think about seriously. But here’s our experience: AI is not replacing people, it’s transforming roles. Our analysts now partner with AI to do more high-value work. AI does the heavy lifting, like data crunching, formatting, parsing. The human sets the direction, interprets output, and adds insight.

We haven't had a single client reduce analyst headcount due to AI. Instead, they're asking those same analysts to help with broader, more strategic initiatives. Think of it as shifting from production to interpretation.

Yes, it changes the skillset, there’s more emphasis on tech fluency, data interpretation, and cross-functional understanding. But it’s a positive shift, and our team embraces it.

With Agent Fleet in place, how do you see Acuity evolving? Will it change your positioning compared to consulting firms?

It definitely changes how we work with clients. Our analysts are now empowered with domain-trained AI, built from years of expertise. While consulting firms are often more generalist, we can go deeper now with  greater precision and impact.