Between the Hype and the Reality
By: Bryson Hamilton, CFA
If you have been following the AI conversation through mainstream media, you would be forgiven for thinking the technology is either about to replace your entire workforce or is an overhyped party trick. The reality, as usual, is far less dramatic. The more useful questions, and the ones business owners are really asking, are simpler: How is AI actually being used in the workplace, is it working, and is it worth the investment?
At NCP, our team has spent the better part of the last year integrating AI into the daily workflows of middle-market investment banking. After thousands of hours of real-world applications, we have developed a more grounded perspective on what this technology can and cannot do.
Some things are better than advertised. Research that once consumed an entire afternoon can now be synthesized and structured in a fraction of the time. Draft documents that once required days of brainstorming and blank-page staring can get a running start in minutes. Raw financial data that used to require hours of manual reworking is consolidated, reconciled, and analyzed at a pace that simply was not possible before. The ability to rapidly iterate analysis by pressure-testing assumptions, reframing arguments, or running alternative scenarios has resulted in more robust data analysis to support negotiating positions.
All of these efficiencies allow our team to focus on the relationship building, judgment, and strategy that set the stage for successful negotiations on behalf of our clients. This is where the real advantage is. Faster iteration, more rigorous analysis, and more thorough scenario testing allows us to use time more efficiently and reallocate time to the judgement of calls and relationship development that drive results.
Other things are exactly as messy as the skeptics suggest. AI does not replace judgment, and it certainly does not replace the relationships, negotiation instincts and industry knowledge that drive middle-market M&A. It hallucinates. It gets confident about things it should not be confident about. But when you understand those limitations and design your workflows accordingly, the technology becomes a real productivity tool instead of a gimmick.
Building the Toolkit
One thing that differentiates NCP’s approach to AI from what you might read about larger organizations is that we are not implementing some enterprise-wide platform mandated by a CTO with three layers removed from the actual work. Our analysts, the employees closest to the work, are the ones driving adoption and deciding which platforms to implement. They are identifying use cases, testing tools, and building workflows tailored to our operational needs, which has resulted in several redesigns of processes that are primed for automation and rapid synthesis.
We use multiple AI platforms, each for what it does best. For analytical heavy lifting, building financial models, structuring presentations, and drafting client deliverables, we have integrated AI deeply into our Excel and PowerPoint workflows. For due diligence and research, we use secure1 AI-powered knowledge management tools that function as living research vaults, where we aggregate and synthesize company data, industry reports, and deal intelligence so our team can query months of accumulated research in real time.
Perhaps more importantly, we have gone beyond off-the-shelf usage. Our team is actively building custom AI workflows that encode our firm’s processes, formatting standards, and analytical frameworks directly into the tools themselves. Instead of asking a generic AI to “build an NCP-branded pitch deck,” our workflows are already tailored for what an NCP-branded pitch deck should look like, what our clients and prospects expect to see, and how our deliverables should be structured. This level of specificity matters because professional services work, M&A in particular, is rarely generic; that means precision, consistency, and the ability to record institutional knowledge are differentiators. AI allows us to build these advantages into our operational infrastructure.
Beyond the Subscription
Here is the part of the AI story that does not get enough airtime: none of this works without the right environment. Technology is table stakes, and any firm can buy a subscription to an AI tool of choice. What separates the firms that will capture true value from AI and those that will not is whether they build a deliberate structure around adoption or just support their employees in identifying use cases that will improve their workflows.
Last year, NCP established an internal AI committee that meets regularly to share new use cases, evaluate data security protocols, and collaborate on what is working and what is not. In an environment where new tools and models are released on a near-daily basis, having a regular forum to share those findings keeps the entire team operating with the most current and effective tools available. For a firm our size, the ability to stay nimble and adapt in real time is a meaningful competitive advantage.
We are still in the early innings of the AI buildout, but the returns are already real. Our team is handling more work2 without adding headcount3 and pursuing opportunities that would have been difficult to execute efficiently only a few years ago. For our clients, that translates to more bandwidth dedicated to the work that matters most and a partner that is building the infrastructure to stay ahead of the curve, not chase it.
AI will almost certainly reshape professional services over the next decade. The firms that benefit most won’t be the ones chasing every headline or newest release. More likely, the winners will be the organizations that integrate these tools into existing expertise, institutional knowledge, and client relationships. In that sense, AI may end up looking less like a technological revolution and more like every meaningful productivity shift before it: gradual, uneven, and ultimately decided by execution rather than excitement.
- NCP exclusively uses enterprise and business-tier AI subscriptions with robust data privacy protections. These platforms isolate our data from public models, meaning our inputs are not used to train the underlying AI systems, information is encrypted in transit and at rest, and access is ring-fenced to our organization. We are deliberate about which platforms we adopt and how client information interacts with them, because confidentiality is foundational to our business. ↩︎
- General estimate from NCP President is +/- 25% productivity enhancement across the team. ↩︎
- Counterintuitively, we believe our future ‘novice’ analyst hires may trend away from 2-3 year required experience toward recent college graduates if and as they are incorporating their “native” experiences in “AI” to the financial disciplines in real-time learned in university. ↩︎
Image created using generative AI