Ben Burge & Matt Miller of Rupp Pfalzgraf LLC
In this episode of the Empire State Entrepreneurs and New York Business Law Podcast, host David Pfalzgraf is joined by two of Rupp Pfalzgraf's own: Matt Miller, Practice Area Leader of the Labor & Employment group, and Ben Burge, M&A attorney and the firm's General Counsel. Together, they pull back the curtain on how Rupp Pfalzgraf has navigated AI adoption, from early experimentation to building out a full internal task force, weekly training sessions, and hiring in-house data engineers to build proprietary tools.The conversation is equal parts honest and practical, touching on what went right, what was scary, and what business owners of any size should be thinking about right now.Recorded live at Incept's podcast studio, The Vault, in Buffalo, NY.Show NotesThe Moment It Clicked – Ben's "Deep Research" WalkBen's entry into AI wasn't strategic; it was accidental. During COVID, he was out for a morning walk with ChatGPT open on his phone, working through a thorny question about who legally owns medical records produced by a physician. He hit the deep research button by mistake. When he got back from the walk, a 25-page, single-spaced memo with citations was waiting for him – work that would have taken him and a couple of associates several days. He admitted he didn't check all the citations (an early lesson in hallucinations), but the experience made clear that something fundamental had changed.The Moment It Clicked – Matt's Hour with Tony RuppMatt traces his AI awakening to a five-minute check-in with Rupp Pfalzgraf managing partner Tony Rupp that turned into an hour-long conversation. Rupp's passion for the topic made it land differently; Matt recognized immediately that this wasn't a passing trend, and walked back to his desk ready to start experimenting.Building the Foundation: Survey First, Strategy SecondBoth Matt and Ben emphasized that Rupp Pfalzgraf's structured approach started with a simple internal survey – asking employees what AI tools they were already using on their own. The results surprised leadership and gave them a real baseline: which platforms were in use, how often, and for what purposes. From there, the firm made deliberate decisions about which tools to standardize on, what enterprise-level accounts to establish (with appropriate confidentiality and security provisions), and how to build a shared strategy from the ground up.The AI Task Force & Weekly Breakfast SessionsRupp Pfalzgraf formed a cross-functional AI Task Force drawing people from all levels of the firm. Regular brainstorming sessions identified inefficiencies where AI could create immediate ROI – particularly finding where high-value employees were spending time on low-value, automatable tasks. The firm also launched weekly Friday breakfast sessions where attorneys and staff shared use cases, explored new tools, and workshopped prompting techniques. Ben led a session specifically on prompt quality, making the case that how you ask matters as much as what you ask.Going All In: Hiring Data EngineersOne of the more surprising moves the firm made was hiring multiple in-house data engineers – a significant departure for a law firm. Rather than simply purchasing licenses for off-the-shelf platforms, Rupp Pfalzgraf saw an opportunity to build proprietary, customized tools tailored to their specific workflows. Current custom builds include automated subpoena generation and document request tools, freeing up attorneys and paralegals from non-billable, repetitive work. The data engineers attend every AI Task Force meeting and take direct requests from lawyers, paralegals, and firm leadership.Measuring ROI – And Why It's the Wrong Question (For Now)When pressed on return on investment, Ben pushed back on the framing. Direct proportional ROI is hard to calculate when the gains are distributed across dozens of people doing their jobs faster and better. The better question, he argued, is whether you're maximizing the highest and best use of each person's time – a phrase David uses often with the firm. The firm has made the conscious decision to invest ahead of measurable returns, betting that productivity gains and revenue upside will compound over time.Quality Control: Accountability Doesn't Come with the OutputA key theme that emerged in both the formal interview and the post-recording discussion: AI gives you an answer, not accountability. Ben and Matt both stressed that the output needs to be owned by the person who generated it. Matt described his approach to reviewing work from junior attorneys – he can tell when AI was used, and rather than penalizing it, he asks pointed questions: Why is this argument here? How does this fit the overall strategy of the case? Did you actually read through what it generated? The goal is to build judgment in younger lawyers, not just efficiency.AI and the Future of RecruitingBoth guests expect the profile of entry-level professionals to shift meaningfully. Real-world work experience before law school is already attractive; Ben predicts it will become more so as AI handles more of the mechanical drafting work – leaving judgment, client relationships, and strategic thinking as the differentiators. Matt noted the challenge facing employers in hiring: writing samples may be AI-generated, and interviewers need to be more probing and observational than ever. He also flagged that AI-based applicant screening tools are a growing area of concern in labor and employment law.Five Years Out: Evolve or DieNeither Matt nor Ben expects AI to replace lawyers in the next five years. But both were clear that the firms and professionals who aren't actively building a strategy right now are already behind. Matt compared the current moment to earlier technology inflection points – computers, email, online legal research – where every professional services industry had to adapt or get left behind. The difference now is the pace; what changed over decades in previous cycles is happening in months.Key HighlightsBen's accidental deep research moment: A 25-page memo on a question that would have taken days, generated during a 20-minute walk. The citation hallucinations were a lesson; the productivity revelation was immediate."AI can give you an answer. It can't give you accountability." – Ben BurgeHiring data engineers at a law firm: A bold move that reflects the firm's ethos of thinking differently and investing in what others won't.Prompt engineering matters: Not a buzzword – the quality of what you put in directly determines the value of what comes out. Ben led an entire internal training session on it.The Tony Rupp BlackBerry analogy: If you're not intentionally investing in something this revolutionary, you're already way behind.Quality control as mentorship: Matt's framework for reviewing AI-assisted work from junior attorneys is really a framework for developing professional judgment in the next generation.The Friday breakfast sessions: A low-lift, high-impact format any business could replicate to build internal AI fluency.Key Takeaways- Start with a survey. Before building any AI strategy, find out what your people are already using. You'll be surprised – and you'll have real data to work with.- Establish security before scale. Enterprise accounts, confidentiality provisions, and data ...




