DANDA #056 — CaseClerk: AI Agents to End Small Claims Court Paperwork Nightmares

Justice isn’t just for the wealthy or business-savvy. Yet, for millions of Americans, small claims court—meant to resolve disputes quickly and affordably—remains a bureaucratic maze. People give up on collecting debts, landlords lose out on damages, tenants suffer unlawful withholdings, and freelancers write off unpaid invoices. The root cause: paperwork nightmares, confusing requirements, and missed deadlines that make the process harder than the dispute itself.

The problem

Every year, over 20 million small claims cases are filed in the US alone. But for every case filed, 1-2 more are abandoned due to paperwork confusion, procedural errors, or missing deadlines. The average American spends 10-15 hours navigating forms, assembling evidence, and deciphering court rules, only to be tripped up by procedural mistakes or missed hearings. Lost claims and dismissed cases cost individuals and small businesses an estimated $13 billion annually in uncollected judgments and unrecovered damages. The courts themselves are swamped: nearly 30% of small claims cases are dismissed for incomplete or incorrect filings, wasting everyone’s time.

The idea: CaseClerk

CaseClerk is an AI-powered agent that acts as a user’s personal clerk for small claims cases. It automates paperwork prep, deadline tracking, evidence collection, and communication with court staff—demystifying each step. Users simply describe their dispute ("My landlord didn’t return my deposit"), upload basic documents, and CaseClerk orchestrates everything: selecting the right forms, pre-filling details, scheduling reminders, and even generating polite, court-ready emails or printouts. For evidence, it creates a visual timeline, extracts key facts, and bundles exhibits—no more mess of screenshots or missing receipts. The agent updates users on case status and court communications, ensuring no deadline or hearing is missed. It’s the justice system—finally made simple, accessible, and human.

Architecture

CaseClerk receives user inputs—plain language, receipts, emails, or photos. The Ingestion layer parses and extracts details via OCR and LLMs, feeding structured facts into a case-specific knowledge graph (timeline, parties, amounts, deadlines). The Agent Orchestrator plans the workflow: forms to prepare, evidence to request, and key dates. Whenever a legal decision or submission is needed, the Human Gate asks for user confirmation or edits. The Action Layer then generates court forms, notices, and reminders, ready for e-filing or printing—keeping the user on track, error-free.

Build plan (90 days)

The wedge: single jurisdiction pilot (e.g., Los Angeles County), focused on tenant-landlord disputes and unpaid invoices. Integrate court form templates, rules, and deadlines into ingestion and agentic flows. Stack: expand to additional case types (auto, consumer, gig work), add integrations with e-filing portals and SMS/email reminders. Pricing: $19/case for DIY, $99/case for white-glove review, with volume rates for legal aid orgs and small business bundles.

Why now

Courts are finally digitizing, and consumers are demanding access to justice without lawyers. GenAI can now parse evidence and optimize workflows with accuracy lawyers charge $200/hr for. The stakes—billions in lost claims, millions frustrated—are immediate. With agentic automation and a human-reviewed action layer, we unlock justice for all, not just the few.