AccessHOU

AccessHOU: prototype interface and core workflow.
AccessHOU: prototype interface and core workflow.
AccessHOU: prototype interface and core workflow.
AccessHOU current live app interface (captured 2026-09-27 19:40 UTC).

Inspiration

The accessibility challenge asked how residents and care teams could find adaptive programs and places with information they could actually use. AccessHOU begins by making separate entrance, interior, bathroom, arrival, sensory, and other access details visible instead of collapsing them into a single label.

What it does

AccessHOU provides a source-backed Houston directory of programs, spaces, and services. Visitors can filter by mobility, sensory, communication, vision, hearing, cognitive, caregiving, and referral needs, compare explicit details and unknowns, save a local shortlist, and use English or Spanish. Authenticated observations enter a review queue.

How I built it

The prototype uses Vite and vanilla browser modules with an Express and TypeScript API, PostgreSQL and Drizzle ORM, Google Identity Services for observations and moderation, and a source-backed seed catalog. Browsing and shortlist state remain local to the device.

Challenges

Access needs are specific and personal, while source pages often leave important details unknown. The data model preserves unknowns, keeps provenance visible, and distinguishes a listing from a certification. Moderation and identity boundaries had to be explicit for shared observations.

Accomplishments

The release provides a working catalog, functional-need filters, separated access fields, source and retrieval dates, local shortlists, English and Spanish paths, and a review boundary for observations. The revised public package also has dedicated pitch and demo surfaces.

What I learned

Accessibility information is more useful when it is granular and honest about gaps. A person's ability to call ahead or share a source can matter as much as a filter result.

What's next

The next step is to expand the catalog with residents and care teams, establish a review process for submitted observations, and test which fields clinicians and families use when planning a visit. No partner or professional endorsement is assumed by the prototype.