- Problem
- People in collections rarely know their rights or how to dispute a bad debt.
- My role
- Founder. Former debt collection PM. I led product, brought on two designers, and built most of it with AI tools.
- What I built
- A free "first aid kit": plain guides, an AI chatbot, and a dispute letter generator. No sign-up.
- Results
- All three tools live and free. Revenue only from optional mailing. Built on three stacks with almost no dev budget.
The problem
Over 70 million Americans have debts in collections. Few know their rights or how to dispute a bad claim:
- 35% of American adults with credit files have debts in collection
- 44% say they feel "scared," "confused," or "helpless" when a collector calls
- Only 1 in 20 disputes a debt, yet up to 59% of debt portfolios contain errors
- The average person spends 45+ hours trying to resolve debt issues
These people aren't trying to dodge real debts. They want clarity, fairness, and a way to fix mistakes. The system profits when they stay confused.
Why me
I worked inside debt collection as a product manager. I saw how powerless consumers feel. People in my community started asking me for help. That made the question simple: what if everyone had the insider knowledge I had?
Key decisions
Free and anonymous, on purpose
Debt relief is full of predatory services. To be trusted, DebtCat had to look nothing like them. So no sign-up, no stored personal data, no subscription, and no payment details upfront. The only paid features are optional mailing services. I gave up early revenue to earn trust.
Action over advice
Most debt products give general tips or push consolidation, which often just moves debt around. I built around three steps that lead somewhere:
- Learn: jargon-free guides to consumer rights
- Ask: an AI chatbot for specific questions
- Act: a generator that writes a legally sound dispute letter
I recruited two talented designers, Etele and Dalma, to shape the product with me. Dalma put the core idea well: "People needed to feel safe before they'd engage." We cut friction wherever we could, even nice-to-have features.
Let the product pick the stack
DebtCat moved through three stacks. WordPress got content online fastest. Once Etele and Dalma joined, Webflow gave us the design control we needed. Firebase cut costs, because we were paying for features we didn't need yet and I wanted that budget for users. Each move cost a weekend of rebuilding and some technical debt.
Build it myself with AI
A few years ago this would have needed a dev team. I planned each feature the usual PM way, then used ChatGPT, Cursor, and Windsurf to draft components, wire them up, and debug. AI didn't replace product thinking. It removed the wait between deciding and shipping.
What got in the way
Distribution. Email providers and Google Ads flagged us for talking about "debt," the same alarms that catch predatory lenders. So we leaned on social media and community work instead of email and ads.
Results
- Guides, chatbot, and dispute letter generator are live at debtcat.com, free with no sign-up
- A clear business model: optional paid mailing, nothing else
- Built and run with almost no development budget
Takeaways
- Domain experience finds the problem. Working in collections showed me both the gap and the fix.
- In a skeptical market, trust comes before revenue.
- Plan distribution early. In regulated-sounding categories, the usual channels can close on you.
What's next
Tools to negotiate directly with collectors, credit report imports to see every debt in one place, and partnerships with reputable credit counseling services. I'm also considering making DebtCat a nonprofit to reach more people.