GoodByeDebt

Over het initiatief / About the initiative

In welke fase zit jouw initiatief? / In what stage is your initiative?

Uitgewerkt conceptfase

Heb je jouw initiatief al gevalideerd? / Did you validate your initiative?

We validated the initial concept through discussions with domain experts, including debt advisors from the municipalities of Leiden and Zoetermeer, and a financial psychologist from the Knowledge Center in Leiden. Currently, we are conducting a structured validation and demo test in collaboration with the guardianship organization Stichting Veritas. As part of this process, we tested our system on 50 real debt letters to assess its ability to extract key information, summarize content, and categorize letter types. The model achieved approximately 98% accuracy, demonstrating strong initial performance. Building on this, we have now entered the second phase of testing, which focuses on combining information extraction with identifying missing fields and automatically contacting creditors to request the missing data. Our goal is to fully validate each step of our roadmap by October. This includes: 1. Extracting and structuring information from debt letters 2. Categorizing the letters based on content and intent 3. Detecting missing or incomplete information 4. Follow up automatically with creditors 5. Processing inbound responses and determining whether action is required 6. Updating the client's system with complete, structured data 7. We are on track to complete this end-to-end validation and deliver a robust, scalable automation pipeline for debt communication workflows.

Meer informatie over jouw initiatief / More info about your initiative

GoodByeDebt is designed to tackle a major bottleneck in debt assistance services: the overwhelming volume and complexity of incoming debt letters. Every day, guardianship organizations, municipalities support services, and debt advisors receive large quantities of letters that are difficult to interpret. These letters often contain legal jargon, inconsistent formatting, and critical information spread across multiple paragraphs or even missing entirely. As a result, staff must spend hours manually reviewing each document, extracting key details, and contacting creditors to fill in the gaps. This is not only time-consuming but also delays the support that vulnerable individuals urgently need.

To address this, GoodByeDebt follows a structured roadmap toward full automation. We’ve already developed the ability to read letters, summarize what the letter is about, and extract standardized metadata such as the sender, due amount, and deadlines. The next steps focus on understanding the full debt position consolidating amounts across multiple letters and identifying whether the debt is recurring, overdue, or disputed. When information is missing, GoodByeDebt will proactively request clarification by contacting the relevant parties or navigating portals where needed without any manual intervention.

Our roadmap also includes processing inbound replies automatically, identifying whether a message requires action or is just informational, and then updating the system accordingly. Eventually, the platform will be able to distinguish which client or case the letter belongs to, allowing us to scale this solution across multiple organizations with minimal friction. Our ultimate goal is to remove administrative burdens from caseworkers so they can focus on what truly matters: supporting people in getting out of debt. Through automation, standardization, and plain language summaries, GoodByeDebt aims to bring clarity, speed, and transparency to the debt resolution process.

Wat is er anders/nieuw aan jouw idee/oplossing t.o.v. bestaande oplossingen? / What is different/new about your idea/solution compared to existing solutions??

GoodByeDebt stands out by offering a fully automated, end-to-end solution for handling debt letters something that current tools and platforms do not address and they are fully manual. Most existing systems focus on document storage or basic case registration, but they leave the most time consuming part interpreting complex, unstructured debt letters to human caseworkers. These letters often contain legal jargon, inconsistent layouts, and missing information, making them difficult to process quickly or accurately.

Our solution is different. GoodByeDebt not only extracts critical information such as sender, amount due, and deadlines, but also summarizes the letter in plain language, categorizes its intent, detects missing data, and even initiates follow-up communication with creditors when needed. It also processes inbound replies and updates internal systems automatically turning a manual, fragmented process into a streamlined, intelligent workflow.

Importantly, our technology is privacy-preserving and can be deployed locally, making it suitable for sensitive environments such as municipal services and guardianship organizations unlike generic, cloud-based AI tools. 

Unlike other systems, we don’t just store documents. We explain, summarize, and help take the next step. It’s not about replacing people, but giving them tools that work with them, not against them.

Wat zijn jouw volgende stappen om het verder te ontwikkelen? / What are your next steps to develop the initiative?

Our immediate next steps focus on completing and validating each phase of our automation roadmap. Currently, we are in the second phase of our pilot with Stichting Veritas, where we are testing the system’s ability not only to extract and categorize information from debt letters, but also to detect missing fields and automatically reach out to creditors to retrieve them.

After this demo, we aim to begin testing with the municipalities of Leiden and Zoetermeer, as well as other debt support organizations, to further scale and improve the system. These pilots will help us validate the tool across different workflows, document types, and operational environments. By working closely with support staff, we can refine the platform’s functionality, ensure compliance with data privacy standards, and adapt it to real-world case management systems.

Our broader goal remains the same: to complete the full automation cycle, from extraction to action, and prepare the platform for scalable deployment across the debt assistance ecosystem.

Wat heb je nodig om (nog meer) impact te maken met dit initiatief? / What do you need to make (more) impact with this initiative?

To make a greater impact, we need strategic partnerships with municipalities, guardianship organizations, and debt support providers. These collaborations are essential not only for scaling the solution but also for ensuring it integrates smoothly into existing workflows and meets the practical needs of professionals on the ground.

We also require support in the form of funding and implementation capacity including time from caseworkers for onboarding and feedback, and resources for technical integration. This will allow us to accelerate testing, deployment, and refinement of the system.

In addition, guidance and visibility within relevant networks are crucial to help us connect with the right experts, policymakers, and stakeholders. This will enable us to align with ongoing initiatives, benefit from domain knowledge, and ensure GoodByeDebt fits into the broader landscape of debt prevention and financial well-being.