Project DANDA

DANDA #031 — LoanAid: AI Agents to Unlock Disability Loan Relief Without Paperwork Hell

AR Akhil Reddy Danda · 21st August, 2026 · 4 min read
DANDA #031 — LoanAid: AI Agents to Unlock Disability Loan Relief Without Paperwork Hell

Every year, hundreds of thousands of Americans become permanently disabled and are eligible to have their federal student loans forgiven. But instead of relief, most are lost in a maze of medical forms, outdated faxes, and confusing government requirements. An AI solution can unlock billions in aid and change lives overnight.

The problem

Today, about 8.2 million Americans receive Social Security Disability Insurance (SSDI) benefits, and over 700,000 of them hold federal student loans. By law, they qualify for Total and Permanent Disability (TPD) discharge, wiping out their debt. The Department of Education, however, still requires a complex application, physician attestations, and back-and-forth with servicers. As a result, 70% of eligible disabled borrowers never complete their TPD applications. The average time to loan relief is over 18 months—many give up or never even try. In total, an estimated $40 billion in eligible loans remain unforgiven, costing disabled Americans over $400 million a year in unnecessary payments they cannot afford.

The idea: LoanAid

LoanAid is an AI-powered agentic platform that automates the TPD discharge journey for disabled borrowers. A user connects their loan information and disability status; the agent gathers required medical evidence, fills out forms, interfaces with loan servicers, and tracks case status—eliminating manual effort. For those with complex cases, LoanAid coordinates with physicians and government portals, handling appeals and follow-ups. The agent uses secure EHR integrations and document extraction to minimize user effort. Agencies, social workers, and legal nonprofits can deploy their own branded versions at scale.

Architecture

User Inputs Loan Info, EHR, Disability Docs Ingestion OCR, EHR API, PDF Parse Case Memory Graph: Status, Docs, Contacts Agent Orchestrator Task Planner, Multi-Agent Routing Human Gate Borrower / Advocate Review Action Layer E-File, Fax, Upload, Portal Login

The user starts by uploading loan details and disability proof. The Ingestion layer pulls in data from PDFs, EHRs, and government APIs, extracting all necessary evidence. The Case Memory graph tracks status, documents, and involved parties. The Agent Orchestrator sequences multi-step tasks—requesting physician attestations, filling forms, and tracking servicer responses. When human review is needed (complex cases, new documents), it routes to the borrower or advocate. The Action Layer files paperwork electronically, faxes to providers, and updates federal portals—automating 90% of the process. The system learns from outcomes to streamline future cases.

Build plan (90 days)

Why now

Loan servicers have finally opened up digital APIs, and EHR interoperability is mainstream, slashing integration time. Regulators are pushing for higher TPD discharge rates, and legal aid groups are desperate for automation. Millions of disabled borrowers need relief urgently—AI can deliver it in months, not years.

in Share on LinkedIn 𝕏 Post
Sources I read for this:
← More from Reddy Pulse