Your best collectors are burning out.
Not because they’re bad at their jobs, but because their collections jobs are bad for them. They’re drowning in a grueling daily cycle of triaging accounts, sending reminder notices, and then manually processing a flood of unpredictable inbound replies.
On average, a collector spends eight minutes on every single email, according to Billtrust’s data. This isn’t eight minutes of focused, high-value work. It’s often a scramble of context-switching: finding the account, re-reading long email threads, manually extracting key details, and then pivoting to another system to log that information. For collectors, it feels like fighting a losing battle against the inbox. For finance department leaders, it feels like being stuck in a war of attrition with a direct-line impact on working capital.
“Our collections team was in a constant state of flux. Every time we hired, we’d lose someone. Growth was our goal, but our processes in place would not allow that to happen.”
Brian Page, Director of Credit, 84 Lumber
Busywork is a recipe for burnout as much as it is a multi-million-dollar crisis lurking in plain sight. Out-of-control collections operations often result in substantial financial write-offs simply because the accounts receivable team can’t keep up. This explains why piles of 8-minute emails can ultimately cost companies thousands — even millions of dollars.
Take a look at bad debt. It hasn’t been moving in the right direction for most of us. Last year, write-offs doubled, and with macroeconomic headwinds picking up, finance leaders are making more defensive moves to protect their financial positions.
Don’t miss the full discussion in our podcast with Launchpod.
The Flaw in Fixing Collections Problems with AI
The common reaction is to “sprinkle a little AI on it,” but technology alone can actually make things worse. Clumsy automation can damage the customer relationships collectors work so hard to build.
The push for AI from the executive level is entirely legitimate, of course. Companies are in a perpetual search for technology that gives employees superhero capabilities. However, while leaders chase a shiny new feature set for their next competitive edge, they often make a big misstep: thinking AI is a magic wand. It’s not.
AI alone is a blunt instrument with the potential to create as many problems as it solves. The challenge isn’t accessing this new technology – it’s turning it into a trustworthy problem solver.
The biggest problem in collections isn’t a lack of technology… it’s a lack of trust in it.
Trusting AI: You’ve Got to Train It
In a sector like finance, where a single mistake can have significant financial and reputational repercussions, the ability to trust the machine’s choices is paramount. At Billtrust, we ask one simple question: “Can you trust my job with it?” If the answer isn’t an immediate “yes,” there’s work to be done.
Collections automation software must be trained by the right data and coached by the right leaders before they can be shaped and molded to handle AR operations and its many scenarios.
That’s why Billtrust built our AR automation platform using a philosophy that we call the “human in the loop.” Our solutions allow direct human oversight, the capacity for verification, and the power to adjust or override machine-generated actions before they are executed. It’s a powerful feedback mechanism to train the AI engine, so it learns from every interaction and gets smarter with time.
“Human-aided automation should always precede unaided automation. AI should be your copilot long before you take the passenger’s seat on the AI plane.”
Dave Ruda, VP of Product, Billtrust
Buyer Beware: Unruly Automation Can Damage Customer Relationships
While some software providers claim their AI platforms have human-aided automation, the reality is most don’t. Providers skip this critical step, which is why people can’t trust their jobs to AI tools. It isn’t getting smarter with every new situation.
AI is like a blunt butter knife. But when guided by human expertise, it can be turned into a scalpel. Billtrust helps you build that AI scalpel – the kind of precision tool you can count on to do your job reliably. After all, unruly automation can damage customer relationships.
Let’s take a closer look at how AI and aided automation are helping collectors tame the chaos of their inbox.
From Crisis to Control: Collections Automation Software that Offers More AI Supervision
In July 2025, Billtrust released groundbreaking new AI capabilities for collections work. Powered by Agentic AI, Billtrust’s new collections agent is a virtual assistant designed to attack those 8-minute email problems. As a collector’s copilot, it:
- Handles the monotonous triage activities so AR teams don’t have to
- Securely scans inbound emails and groups them into easy-to-follow categories, such as instantly categorize such as ‘promise to pay’ or ‘disputes’
- Extracts key data and adds it to the email’s summary, together with any important information form the email thread
- Makes next step recommendations, including drafting response emails with all the right client details and files attached
- Creates tasks for the collector based on the email thread and action needed
- Drafts response emails with all the right client details and files attached, which the collector can edit or adjust to match their tone of voice
- Uses aided automation, combining the initial AI-generated responses with the collector’s previous edits to develop a “voice of the collector” — this ensures that all future communications feel natural and personalized, not robotic
This information is then presented to the collector as a simple, one-click action for verification, giving the collector the confidence that AI is carrying out the process correctly. This one click is arguably the most important, because it offers a system of checks and balances that empower the collector to coach the AI engine to behave the way it should.
Real-World Collections Results: Reducing 8-Minute Emails to 2.5 Minutes
Billtrust’s clients have recognized immediate and dramatic results. By focusing on the real bottleneck—that eight-minute crisis—they’ve cut their email handling time to just 2.5 minutes. That’s a 3.2X increase in capacity! Don’t miss the story of Peak Industrial, which made a $300,000 dent in their bad debt, and listen in as 84 Lumber explains how they solved their collections problems.
Clients see a 3.2X increase in capacity with Billtrust’s Agentic Email features.
Rescuing People from Remedial Work
More than just efficiency, embracing a trustworthy AI tool is about giving your team the breathing room to do the work that humans are inherently good at—connecting with clients through meaningful conversations, negotiating with them around payments, and solving complex problems. When trained correctly, AI can dismiss teams from the drudgery of their inbox, empowering them to engage in more accounts, resolve disputes faster, and have a direct, measurable impact on corporate cash flow.
The end result isn’t jobs eliminated, but rather jobs elevated.
By automating the mundane, Billtrust aims to increase job satisfaction and retention, enabling our clients and their employees to focus on tasks that contribute to their organization’s mission. And our “human in the loop” philosophy extends beyond the collector’s inbox, it’s informing how we use AI to streamline everything from dispute management to credit analysis.
Learning Loops are a Pathway to AI Autonomy
In looking ahead, Billtrust is integrating our “human in the loop” philosophy into our broader solution design. Beyond just a feedback loop, we’re developing learning loops where AI continually learns from every edit and correction.
While the current focus is on refining AI engines with learning loops, the ultimate vision is a future where AI can be trusted to handle low-risk, high-reward tasks autonomously. The end goal is to establish a new baseline where manual burdens are lifted by robotic process automation. One day, learning loops will free finance teams to focus their energy exclusively on high-value work – the work they were actually hired to perform.
Historical lessons learned have already underscored the need for this more thoughtful approach to AI innovation — where human-aided automation lays the groundwork for full autonomy.
We believe the next decade will be defined not by which companies adopted AI, but by which ones have thoughtfully adopted it.
At the end of the day, a true AI strategy isn’t about chasing trends or having the biggest number of AI agents in your toolkit. It’s about supervision and shaping these next-gen technologies into the smartest solution for your business.
Ready to learn more about how AI is revolutionizing collections functions? Don’t miss the full podcast discussion, ask for a personalized product demonstration of Billtrust’s Collections solution.
