The AI Agent Behind 4x Productivity in Collections Teams | Dave Ruda, VP Product (Billtrust)
From cutting email time by 70% to designing human-in-the-loop AI for trust and adoption, Dave Ruda shares how Billtrust built an AI collections agent that delivers real ROI for collections teams.
TL;DR
In this episode of LaunchPod, we sit down with Dave Ruda, VP of Product at Billtrust, to explore how his team built an AI collections agent that boosted productivity 4x for finance teams.
Dave breaks down how to spot real opportunities for AI (beyond “Salt Bae-ing” it on top of features), why human-in-the-loop is essential in FinTech, and how Billtrust is designing AI tools that collectors actually trust and adopt.
Here’s what we cover:
How Billtrust built an agent that cut email response time from 8 minutes to 2.5 minutes and scaled capacity for thousands of collectors (03:00)
Why product leaders must go beyond buzzwords to find real AI use cases that solve customer pain (06:00)
How human-in-the-loop safeguards protect trust in high-stakes finance workflows (09:00)
Why adoption depends on designing tools that fold naturally into existing workflows (20:00)
1. Building an AI Collections Agent That Delivers Results
Instead of just adding chatbots, Billtrust focused on a real pain point: collections emails. Collectors typically spent eight minutes parsing each inbound email. By deploying an AI agent that categorizes responses (promise to pay, dispute, missing invoice) and surfaces them as single-click actions, Billtrust cut that time to just 2.5 minutes.
The result? A 4x productivity gain and thousands of additional accounts touched without extra headcount.
Things you can do
Audit your customer workflows to find repetitive, high-volume pain points (like parsing inbound emails)
Prioritize AI use cases where time savings can be clearly measured in minutes, not just “nice-to-have” improvements
Start with a single high-friction task instead of spreading AI across your entire product
2. Finding Real AI Opportunities in Customer Pain
Dave emphasizes that the right AI strategy starts with customer pain, not hype:
“The real question is how you can address some of the pain points in the process where customers see this pain and then be able to address it — because the technology now exists.”
For Billtrust, the bottleneck wasn’t sending automated reminders; it was processing the messy, unstructured responses. By applying GPT models to categorize and log outcomes, collectors could move faster while focusing on the highest-value accounts.
Things you can do
Interview frontline users to map where time is wasted or bottlenecks occur
Frame AI not as a feature but as a pain reliever. Show exactly what task it eliminates
Run quick tests on small datasets to validate whether AI can meaningfully reduce effort before scaling
3. Human-in-the-Loop Is Essential in FinTech AI
In FinTech, trust is everything. Automating too aggressively risks mistakes with massive financial consequences. That’s why Billtrust designed their agent around human-in-the-loop validation:
“The question I always ask myself when I’m using a new tool is, do I trust my job with it? Because if I’m wrong, maybe I don’t have a job anymore. We work in finance — you’ve got to be able to trust that you’re making the good choices.”
Things you can do
Communicate clearly where AI is assisting versus where humans are accountable
Use mistakes as signals: If users override AI often, refine your model or narrow its scope
4. Adoption Depends on Workflow Integration
AI adoption often fails when tools feel bolted on. Billtrust instead designed their agent to fit seamlessly into existing workflows.
By pulling unstructured email data into structured actions within collectors’ existing tools, the agent feels like a natural extension of their daily flow. This not only drives measurable ROI but also improves retention by removing drudgery and allowing collectors to focus on meaningful conversations.
“What we see is like, people go, oh my God, are they gonna replace me? That’s not the answer. The answer is we just want you to have more meaningful conversations where they need to happen.”
Things you can do
Keep AI invisible: The less users feel like they’re “switching modes,” the higher the adoption will be
Frame AI as a partner that removes drudgery, not a threat that replaces roles
Chapters
00:00 Introduction
03:18 Billtrust's AI Journey
05:10 "Human in the Loop"
06:22 Billtrust's FinTech Solutions
09:54 Challenges in Collections
13:04 Agentic Email Functionality and Automating Email Processing
17:19 Building a Learning Loop To Better Understand Customers
19:05 Empowering Product Managers With AI
24:33 AI Governance and Future Prospects
27:39 Concluding Thoughts and Future Plans
Links
What does LogRocket do?
LogRocket's Galileo AI watches user sessions for you and surfaces the technical and usability issues holding back your web and mobile apps. Understand where your users are struggling by trying it for free at LogRocket.com.