Automate AMI Income Sorting for Housing Applications

Most leasing teams start by manually reviewing PDF applications and using a calculator. This breaks during a lease-up when hundreds of applications arrive at once. A single agent trying to annualize wages (hours x 2080), tips, and bonuses from messy paystubs will make errors. A mistake placing a 60% AMI applicant in a 50% AMI unit creates a significant compliance violation found during audits.

Standard property management software like RealPage or AppFolio are systems of record, not automation engines. They have fields for AMI status, but they do not automatically parse documents, calculate anticipated income, and sort the waitlist. The workflow remains manual: download PDF, open calculator, type number into a spreadsheet, then manually update the record in the PMS. This multi-step process for every single applicant is the source of the bottleneck.

The core problem is that generic document parsers or basic workflow tools cannot handle the specific logic of income anticipation for affordable housing. They cannot distinguish trailing 12-month income from projected income or apply asset limitation tests for HOME-layered units. This forces teams into manual work that does not scale past a few dozen applications per week.

Syntora's approach to automating AMI bucket sorting involves a structured engagement focused on your specific compliance needs and technical environment.

We would begin with a detailed discovery phase to audit your existing application workflow, identify all relevant income documentation, and precisely define the LIHTC, HOME, or HUD program rules and calculation methodologies applicable to your properties. This includes understanding variable income projections, asset tests, and specific exemption criteria.

The technical architecture would center on a custom-built, API-driven backend. We would develop integrations with your RealPage or AppFolio API to automatically ingest new application data. For income verification, the system would utilize the Claude API to parse uploaded paystubs, offer letters, and bank statements. Syntora has extensive experience building robust document processing pipelines using Claude API for sensitive financial documents, and these same proven patterns would be applied here to extract granular income details from diverse document types.

A core FastAPI service, written in Python, would contain your specific compliance logic. This service would process the parsed income data, apply annualization standards (e.g., 2080-hour rule), project variable income, and conduct asset checks. It would then accurately assign each applicant to the correct AMI bucket (30%, 40%, 50%, 60%, 70%, 80%). The design goal is a highly efficient and auditable calculation engine.

The final AMI determination would be written back to a designated field within your RealPage or AppFolio system. We could also implement a separate Supabase database to maintain a dynamic, tiered waitlist view for your leasing team, providing real-time insights into qualified applicants. Supporting components, such as automated applicant confirmation emails via AWS Lambda and robust system monitoring with AWS CloudWatch and structlog, would be integrated to ensure reliability.

A typical engagement for developing and deploying this type of custom automation solution takes approximately 10 to 16 weeks. Key client contributions would include access to relevant APIs, comprehensive documentation of your program rules, and anonymized sample income documents for system training and validation. The deliverables would comprise the fully functional, custom-engineered system deployed within your cloud environment, complete technical documentation, and hands-on knowledge transfer for your operational and IT teams.