RealDoor

The description states that RealDoor is a tool for renters applying to affordable (LIHTC) housing. It allows users to upload pay stubs (PDF or photo), extracts only allowlisted fields with confidence levels, and compares confirmed income against HUD MTSP income limits for their household size. It then assembles a downloadable application packet without making eligibility determinations.

• Product function: Income verification and application packet assembly.
• Core features: Pay stub extraction, income comparison to HUD thresholds, checklist status, and clean packet generation.
• User journey: Three-step process: Profile → Understand → Prepare.
• Data handling: Session cookie-based, no login; data deletion possible with one click.

Inference The product is a frontend-to-backend system built for a specific use case in affordable housing. It uses RAG (Retrieval-Augmented Generation) to cite sources but avoids LLM arithmetic or decision-making.

The description states that RealDoor positions itself as a copilot for renters applying to affordable housing, aiming to close the gap between public rules and user understanding. It explicitly rejects being a "black-box AI that decides your future" tool.

• Core positioning: A transparency and readiness tool, not an eligibility decision engine.
• Key claim: “RealDoor shows you the numbers, it never renders a verdict.”
• Evolution of claims: The project evolved from a hackathon idea into a system with strong guardrails around AI use and data handling.

Inference The positioning is rooted in trust and clarity. It avoids making eligibility claims to maintain credibility and avoid legal or ethical complications.

The description states that RealDoor targets renters applying to affordable (LIHTC) housing, particularly those who struggle with paperwork due to lack of clarity on rules or errors in submission.

• Primary customer: Renters applying for LIHTC housing.
• ICP focus: Individuals with low income and limited access to housing resources.
• User profile: Likely not tech-savvy, but motivated by need for accurate application preparation.

Inference The ICP is defined by the specific housing market segment (LIHTC) and user pain points around documentation errors. No evidence of broader targeting or segmentation beyond this.

The description states that RealDoor does not have a login, no pricing model, and no monetization strategy described.

• Business model: Not evidenced.
• Pricing: Not evidenced.
• Monetization: Not evidenced.

Inference The project is a hackathon submission with no business model or pricing structure. It may be intended as a prototype or proof-of-concept.

The description provides details on the tech stack and architecture:

• Backend: FastAPI, Postgres (encrypted), ChromaDB for rules corpus, OpenAI for extraction.
• Frontend: Next.js + shadcn/ui (Radix underneath).
• Data handling: HUD MTSP 2026 income limits; synthetic pay stubs used for testing.
• Deployment: Frontend on Vercel, backend on Render.

Inference The system is built with a focus on security and accessibility. It uses deterministic logic for calculations and avoids LLM arithmetic to ensure trustworthiness.

The description does not state any traction, revenue, or adoption beyond the hackathon submission.

• Traction: Not evidenced.
• Customers: Not evidenced.
• Maturity: Not evidenced.
• Usage metrics: Not evidenced.

Inference The project is a prototype with no evidence of real-world usage or growth. It was built for demonstration and testing, not deployment.

The description does not mention any competitors or market context beyond the general problem of affordable housing application errors.

• Competitive landscape: Not evidenced.
• Market context: Not evidenced.
• Differentiation: Not evidenced.

Inference No competitive analysis is provided. The project appears to be solving a niche problem without known direct competitors in its current form.

The description raises several concerns:

• No revenue or traction: The product is not monetized or used beyond the hackathon.
• No customer data or feedback: No evidence of real users or user testing.
• AI guardrails are self-imposed: Trust depends on internal constraints, not external validation.
• Limited scope: Focuses only on one housing type (LIHTC) and one geographic region (Boston-Cambridge).
• No scalability claims: The system is described as a prototype, not a scalable solution.

Inference The project lacks commercial viability or traction. It may be a useful prototype but not a business.

1. What is the plan for scaling beyond the hackathon prototype?
2. Are there any real-world users or partnerships in the affordable housing space?
3. How does the team intend to monetize or sustain this product?
4. Has the system been tested with actual HUD income limits or only synthetic data?
5. What are the legal and ethical implications of not making eligibility decisions?