Benny the Benefits Navigator

An AI-enabled platform that streamlines Social Security Disability Insurance (SSDI) and Supplemental Security Income (SSI) applications, to reduce completion time and improve accuracy, completeness, and approval rates.
Project Description
Benny is a purpose-built, AI-enabled system for legal aid groups, social service providers, and disability advocates to help clients prepare and submit approval-ready SSDI/SSI applications directly to the Social Security Administration (SSA).
It aims to transform the traditionally fragmented and error-prone application process into a guided, narrative-first workflow that integrates medical evidence, performs automated quality checks, and submits final applications electronically.
Value of the Tool
For Service Providers
The goal is to dramatically reduce the staff time needed for each application—from many hours to under an hour—by streamlining repetitive steps. Workflows aim to improve the consistency and completeness of submitted applications through built-in automated checks and review prompts. Another goal is to enable partial self-service for applicants, allowing many to complete much of the process independently so advocates can focus on complex or high-need cases. The system is designed to connect directly to medical record systems, removing the need for manual document gathering and freeing staff for higher-value work.
For Applicants
The vision is to replace long, static forms with an accessible, conversational intake process that draws out the narratives and details that strengthen an application. By embedding expert-backed best practices directly into the process, the tool aims to increase approval rates and reduce the need for re-submissions.
Applicants should be able to move smoothly from intake to submission without navigating the Social Security Administration’s complex websites or paper forms, making the process faster, less confusing, and more equitable.
How It Was Built
- Design Priorities: Simplicity and operational efficiency, avoiding over-engineering while meeting SSA submission requirements.
- Tech Stack:
- Frontend/UI: HTML, Bootstrap, Vanilla JavaScript for accessibility and ease of use
- Backend: Flask Forms (Python) with OpenAI / Anthropic API calls for AI-driven text capture, validation, and narrative drafting
- Data Layer: SQL database storing applicant and case information
- Automation: Robotic Process Automation (RPA) to transfer data directly from the database to SSA’s systems
- Integrations: Fasten Health EHR APIs for one-click retrieval of medical and provider records
- AI Approach: Focused use of small, targeted AI agents with optimized prompts, rather than heavy fine-tuning or retrieval-augmented generation, to reach 90% of functional goals in early builds.
How It Might Be Adapted
- Beyond SSDI/SSI: Extend to other benefits applications (Medicaid, SNAP, housing subsidies) that require detailed narratives and document uploads.
- For Legal Aid Networks: Integrate into statewide or national platforms to unify intake and case prep workflows.
- Multilingual and Accessibility Enhancements: Adapt prompts, UI, and document generation for multiple languages and accessibility compliance.
- Integration with Justice Data Platforms: Connect with case management systems for automated status tracking, appeals preparation, and reporting.
- Modular Deployment: Components (narrative capture, EHR integration, RPA submission) can be licensed or integrated separately into other justice-tech tools.