In a recent episode of the Monitor podcast, Editor-in-Chief Rita Garwood sat down with Aaron Case, President of TFS Solutions, to explore how his company is applying AI and automation across its operations. Through a practical lens, Case shares how TFS identifies inefficiencies, empowers its team to challenge the status quo and partners with tech providers to drive measurable improvements in speed, accuracy and client satisfaction.
Q: Aaron, can you start by telling us a little bit about yourself and your business?
A: Sure. TFS creates captive finance solutions and virtual F&I programs for OEMs, dealers and marketplaces. We’ve been around for 40 years, working primarily in equipment and auto finance.
Q: What initially sparked your interest in using AI? Were there specific problems you were trying to solve?
A: Yes, it started with what we call our “stupid policy” rule. We pay staff bounties for identifying inefficient or redundant processes. It was frontline staff saying, “I do this same task all day — there’s got to be a better way.” That pushed us to explore automation and AI.
Q: What were the first tasks your team flagged for improvement?
A: The first big one was processing incoming applications. We receive thousands of apps monthly in various formats, including PDFs, Word documents and emails. It was a lot of unstructured data. We had a team to enter that data into our system. It was costly and caused delays, especially in peak season. Our first AI project automated that data ingestion using tools like Oculus.
Q: How did that new process work?
A: We set up a bot behind our email system. When an email comes from a registered partner, the bot scrapes attachments — including PDFs and Word files — reads them and maps key data, such as business addresses or serial numbers, directly into our system. It’s now a self-sustaining process.
Q: Has this automation improved your clients’ experience too?
A: Definitely. It’s about shaving time off at every stage — 20 minutes here, 30 minutes there. We aim to get back to partners in under 4 hours, sometimes 30 minutes, depending on the asset class. Faster turnaround means better service, stronger relationships and higher NPS scores.
Q: Beyond document handling, what other areas have you improved with AI?
A: We’ve built an algorithm for routing credit applications to the best-fit lender based on criteria like industry, credit score, time in business and asset class. Instead of relying on junior analysts, AI helps us route applications instantly.
Q: Are you using AI for underwriting as well?
A: We’re starting to. It’s the riskiest area, so we approached it last. We’re using AI now as a second set of eyes — comparing its recommendation to our chief risk officer’s. It’s not yet smarter than 40 years of human experience, but it’s evolving fast.
Q: How have your lending partners responded to all this automation?
A: Positively. When we send them more of the deals they actually want, their approval rates and efficiency go up. That leads to better pricing for us, which we pass on to our customers. Everyone wins.
Q: What about robotic process automation (RPA)? How are you using that?
A: RPA is great for tasks that involve checking, testing and verifying, like confirming identity, VINs or inspecting bank statements for inconsistencies. AI can verify the info, and RPA completes the task, like generating documents.
Q: Is your tech stack built internally or outsourced?
A: It’s a mix. We have in-house tech capability through a company we co-own in auto finance. But for AI, we work with external vendors — especially those focused on R&D. They’re usually the best fit for cutting-edge solutions.
Q: Are there areas where AI just doesn’t work and the human touch is essential?
A: Absolutely. Sales and relationship management require human connection. Also, for complex credit decisions — especially in non-prime lending — there’s still a lot of nuance that AI can’t fully replicate.
Q: How did you measure the impact of these changes?
A: Before AI, it took about 45 minutes to get an application into our system. Now it takes 60 seconds. That alone made a team redundant — but we didn’t let them go. We reallocated them to areas where they could drive revenue or cut more costs.
Q: How did your team respond to those changes?
A: Very positively. We have a young, dynamic team and a core philosophy we call the EASY principles: Entrepreneurial, Accountable, Simplistic Thinking and Youthful Thinking. They embrace change and want to improve their own workflows.
Q: Any surprise benefits from these improvements?
A: Employee retention. We built what I call a “one screen mentality.” Salespeople can do everything from one screen — push deals, view credit decisions, update notes and email customers. Simpler systems reduce training time and increase satisfaction.
Q: Can you walk us through a borrower experience that’s improved thanks to these technologies?
A: Take auction finance. If a buyer calls the day before an event, we can now get them approved and funded in time. Five years ago, with paper processes, that would’ve been nearly impossible. Now, speed and reliability help us win more business.
Q: What’s next for TFS in AI and automation?
A: We’re always looking. I take every vendor meeting. I don’t claim to be an AI expert, but I’m here to learn. If someone can wow us, we’ll explore how to implement it.
Q: What advice would you give to companies just starting their AI journey?
A: Know the problem you’re solving before you talk to vendors. A lot of them will try to sell you something out of the box. Start with the lowest-hanging fruit — simple, fixable processes — and go from there.
Q: What makes a tech partner stand out to you?
A: The best vendors take time to understand our business and our workflows before pitching anything. Off-the-shelf solutions can miss the mark. It’s essential that they ask the right questions and help identify genuine needs.
Q: Any final thoughts or lessons from your AI journey?
A: Just that I’m always open to sharing what we’ve learned. If anyone’s interested in captive finance or in learning more about how we use technology, I’m happy to chat.
