New-grad veterinarian turns AI builder as vet med tools mature

CURRENT FULL VERSION: Jason Szumski, DVM, is part of a small but increasingly visible group of early-career veterinarians moving from end users of AI to builders of it. In coverage tied to Vet Life Reimagined and reinforced by University of Illinois reporting, Szumski’s story centers on a familiar tension in practice: the jump from seeing one or two cases a day in training to managing a full clinical schedule while still trying to document thoroughly, communicate clearly with pet parents, and build confidence as a new doctor. That experience helped shape VetSOAP, the AI documentation company he co-founded with Aaron Smiley, DVM. (vetmed.illinois.edu)

The background matters. According to the University of Illinois College of Veterinary Medicine, Smiley first approached Szumski about the idea while Szumski was still a student, but the two did not begin building the product until late 2023, after Szumski had entered practice and had a clearer view of what recent graduates were missing. In that account, Szumski framed the biggest gap not simply as medical knowledge, but as confidence, especially for clinicians who may not have strong mentorship structures around them every day. (vetmed.illinois.edu)

That helps explain why this story fits the education-workforce category as much as the technology beat. VetSOAP is described by its founders as a tool that automates SOAP-note creation from audio recordings, but Szumski has also said the software was shaped by the transition challenges he faced as a new veterinarian. University of Illinois reporting noted that the platform was built to search a curated dataset for relevant articles and offer diagnostic support, while AAHA’s 2024 roundup positioned VetSOAP among a broader class of veterinary AI tools aimed at reducing manual documentation and reclaiming clinical time. (vetmed.illinois.edu)

The larger industry context suggests his message is landing in a profession that is interested in AI, but still figuring out how to use it responsibly. A 2025 survey cited in Frontiers in Veterinary Science found that familiarity with AI was associated with greater optimism and adoption among veterinary professionals, and AAHA reported in 2024 that nearly 40% of surveyed veterinary professionals said they were already using AI tools or software in their veterinary setting. Meanwhile, a newer JAVMA-indexed survey published in early 2026 found veterinary workers reported low knowledge of AI overall, even as attitudes toward adoption were generally positive. Taken together, those findings suggest a field that is moving ahead faster than formal education and governance structures may be keeping up. (public-pages-files-2025.frontiersin.org)

Industry reaction has become more practical, and more cautious, than it was a year ago. In a recent Cone of Shame episode, Andy Roark and Aaron Massecar discussed AI scribes through the lens of evidence from human medicine, explicitly arguing that practices should not rely on vendor messaging alone. Roark was also clear that he likes the technology and has found that AI scribes make his days in practice better by taking away much of the administrative work he likes least. But the conversation focused on the harder questions: what these systems measurably improve, what kinds of errors they may introduce, and how privacy, oversight, and quality control are being managed as they move into routine workflow. Roark made a similar point in a separate discussion on ethical use and quality control, noting how frequently AI scribe questions were coming up in conference conversations. That tone matches what many veterinary teams are now wrestling with: not whether AI will be present, but where it belongs in workflow, training, and oversight. (drandyroark.com)

That broader framing also shows up outside the AI-scribe conversation. In Vet Life Reimagined, Christie Long, DVM, described bringing a software and business background into veterinary medicine and now helping rethink what sustainable, high-quality care can look like at Modern Animal for both clients and care teams. In a separate episode, Mike Mossop, DVM, framed AI as a likely “co-pilot” for the profession and argued that technology should enhance, not replace, relationship-centered care. Those conversations do not focus on VetSOAP specifically, but they help explain why tools like it are getting attention: veterinary leaders are increasingly talking about innovation not as novelty, but as infrastructure for safer, more sustainable care delivery.

There is also a parallel conversation happening around training. In another recent Cone of Shame episode, Roark’s discussion with AAVMC Spectrum of Care Initiative leaders centered on preparing veterinarians to achieve “golden outcomes, not a gold standard” and to meet clients where they are, rather than defaulting to referral-heavy or idealized care models that may be out of reach. That matters here because documentation tools, decision support, and workflow automation are only useful if they fit the realities of general practice, variable client resources, and the day-to-day demands placed on early-career clinicians. In other words, doing more with less is not just a staffing problem; it is also a design problem.

Why it matters: For veterinary professionals, Szumski’s story is a useful signal that workforce pressure is now directly shaping product development. Documentation burden, uneven mentorship, and the need to help new graduates practice efficiently without feeling isolated are all longstanding issues. Research indexed by PubMed and cited in recent AVMA materials has suggested that efficiency gains could reduce staffing pressure in companion animal practice, and AVMA has separately pointed to technology, AI, and better record-keeping as factors that could improve service delivery over time. If AI tools are increasingly being designed by clinicians who have just lived the new-grad experience, they may be better positioned to address real friction points. But the surrounding conversation in veterinary media is a reminder that usefulness alone is not enough: these tools also need to support sustainable care models, preserve the human relationship at the center of practice, and be validated in the messy real-world settings where spectrum-of-care decisions, privacy concerns, and supervision gaps all intersect. (pubmed.ncbi.nlm.nih.gov)

There’s also a cultural point here. Veterinary medicine has often celebrated nontraditional careers in hindsight, after founders, executives, or educators have already built something successful. Szumski’s profile suggests that entrepreneurial experimentation is moving closer to the start of the veterinary career path, not the end of it. That idea also fits the wider storytelling now surfacing in veterinary media, where leaders like Long and Mossop are being featured not just for clinical work, but for bringing systems thinking, technology fluency, and alternative career experience into the profession’s mainstream conversation. For students, interns, associates, and practice leaders, that could gradually reshape what “career development” means in the profession, especially as AI, software design, and workflow improvement become more relevant to everyday care delivery. This is an inference based on the trajectory of recent conference programming, trade coverage, and the way AI-focused veterinary startups are now being featured in mainstream professional channels. (connect.navc.com)

What to watch: The next test is whether AI documentation and decision-support tools can move beyond early enthusiasm and show measurable gains in time savings, note quality, onboarding support, and clinician wellbeing, while also satisfying concerns about privacy, accuracy, and clinical oversight. Expect more scrutiny from conference stages, podcast discussions, and practice managers as adoption expands through 2026, especially as the profession pushes to define how AI can function as a practical co-pilot without weakening judgment, mentorship, or client trust. (drandyroark.com)