In episode 143 of The Woodard Report Podcast, we talked with Jeff Seibert, CEO of Digits. Jeff has spent most of his life building technology that helps people work smarter. “I taught myself when I was 13,” he recalled. “My mom randomly gave me the book Mac Programming for Dummies for Christmas that year. I did the example, ‘hello world’. And then on a whim, I decided to make that text turn orange. And the moment it turned orange on my screen, this light bulb went off and I literally ran outside screaming, ‘I can make our computer do anything I ever wanted.”
That moment set the stage for a career in software innovation. Seibert co-founded Crashlytics, which was acquired by Twitter, and Increo, which Box acquired. His current company, Digits, brings real-time, AI-powered accounting to businesses.
Artificial intelligence is everywhere in the headlines these days, and with it comes plenty of confusion. Many companies are quick to call their products “AI-powered,” even when the technology behind them doesn’t truly measure up. Seibert acknowledged the hype but emphasized that the innovation driving real AI is genuine and transformative for the accounting profession.
Digits launched in 2018, long before “AI accounting” became the phrase of the moment. From the beginning, Seibert and his team weren’t chasing buzzwords. They focused on reimagining what accounting could look like. “I felt there was an opportunity to sort of reimagine accounting in a world of machine learning.”
Seibert sees a key difference between the generative AI behind tools like ChatGPT and the predictive models used in accounting. “There’s generative AI, which everyone’s very familiar with - ChatGPT,” he said. “They don’t really have any built-in fact-checking ability and they still hallucinate wildly.”
Digits takes a different approach. “Accounting is not a generative field in general. You don’t want to hallucinate transactions,” Seibert said. “These models cannot hallucinate by nature. They make a prediction on a fixed set of outcomes.”
That distinction matters. Generative models create content based on probability. They complete sentences with the most likely words. Predictive models, by contrast, work from a limited set of defined options. In accounting, that means a model can only categorize transactions using a company’s existing chart of accounts. “It can't invent a new one because that's not in the set you gave them.”
Even as accuracy improves, accountants remain essential. Seibert explained that while predictive models can handle most transactions with impressive precision, they still need human oversight. Digits, for example, only books transactions when the model is highly confident in its prediction and flags anything uncertain for review. “We are nowhere close to a world where AI is perfect,” he said.
That partnership between technology and professional judgment, Seibert added, is where firms will gain their edge. “You are not going to be replaced by an AI,” he said. “You will be outcompeted by a firm that adopts it very effectively.”
Seibert believes the real barrier for many legacy accounting systems isn’t AI, but data architecture. “They’re facing an architecture issue where their data quality isn’t good enough,” he said. “The quality of the data you give the AI is paramount.”
Digits uses a different foundation. “We are not a relational database,” he explained. “We use a vector graph data model and you can think of everything as an object. Uber’s an object, Airbnb’s an object, your travel expense category is an object. That allows these models to basically build a semantic knowledge graph of your finances and really understand the flow of money.”
For Seibert, the pace of change is accelerating. “What’s interesting about AI is it’s happening so much faster,” he said. “It’s a crazy world, it’s moving really, really fast. I can’t even wait to see where it will be 12 months from now.”
Still, his advice to accountants is grounded: start small. “Just start trying everything that’s out there and understand what helps you, what doesn’t,” he said.
As the technology continues to evolve, Seibert believes the firms that experiment early will be the ones best prepared to adapt, grow, and lead through the next wave of change.
🎧 Listen to the full episode at woodard.com/podcast.
Enjoyed this episode? Subscribe to The Woodard Report Podcast on your favorite platform for weekly insights that move your practice—and your life—forward.
Thank you to our show sponsor, Canopy!
Unclunk your firm with Canopy, the fully integrated practice management that helps accountants build the firm they always wanted. The suite includes client and document management, workflow, time and billing, engagements and proposals, and more. Check out getcanopy.com.
This article was written with the assistance of AI and edited by a human.