Inspiration
Debt collection is broken—an enormous, inefficient market plagued by outdated, manual processes. Today’s systems force collectors into repetitive tasks such as account consolidation, segmentation, campaign creation, outbound calling, invoice generation, and reporting. After building a CRM for debt collectors and witnessing persistent human errors, I realized that AI agents could automate these tasks, reduce mistakes, and transform the entire debt collection value chain.
What it does
Accord is an AI agentic platform that transforms outbound debt collection. It uses a voice agent to engage debtors in personalized, compliant conversations that boost recoveries and cut costs—while treating customers with fairness and empathy.
How we built it
Voice Orchestration: ElevenLabs Conversational AI Hosting: Vercel Outbound Calls: Twilio LLMs: GPT-4o
Challenges we ran into
Real-Time Call Orchestration: Integrating ElevenLabs for live voice interactions required creative polling for conversation IDs. Feature Scope: We aimed to include robust debtor segmentation, but time constraints kept our MVP focused on core voice interactions.
Accomplishments that we're proud of
We successfully deployed a voice AI agent that autonomously initiates outbound calls, streamlines compliance, and automates repetitive tasks—proving that automation can significantly improve debt collection outcomes.
What we learned
Time constraints prevented us from building a complete payment portal. However, leveraging platforms like Lovable demonstrated that integrated payment solutions can be developed quickly.
What's next for Accord AI
Our next step is to engage directly with debt collectors and industry experts to deeply understand the workflow challenges. We plan to partner with collection agencies to deploy Accord, refine its capabilities, and build a full-scale solution that transform debt collection.
Built With
- elevenlabs
- lovable
- nextjs
- openai
- shadcn
- vercel
Log in or sign up for Devpost to join the conversation.