All selected work

Showplace / Zero to one · AI video

From a listing link to a narrated property tour.

An AI virtual tour generator for real estate agents. Built and launched as a solo builder, then grown to $10K MRR.

My roleSolo product builder · via Buildroom.ai
ContextReal estate · AI video generation
ShowplaceAI PROPERTY VIDEO
tryshowplace.com
Showplace real estate video product website
The product I built and launched

The challenge

The problem to solve.

Give real estate agents a straightforward way to create narrated video tours from property listings.

My contribution

What I built and owned.

  • Designed the product around a direct workflow: a listing URL goes in and a narrated video tour comes out.
  • Built the video and narration workflow with ElevenLabs and Veo 3.
  • Owned the product from build through launch and revenue growth as a solo builder.

The outcome

What came of it.

Showplace grew from $0 to $10K in monthly recurring revenue. The project brought product design, AI integration, engineering, and launch into one end-to-end build.

$10KMRR
1solo builder
URL → videocore workflow

My ownership

Where I carried the work.

I built and launched Showplace as a solo product builder through Buildroom.ai: the listing-to-video experience, the application and the generation workflow.

The wider team & tools

The product uses external services for property data, narration, generation and infrastructure. My work brings those capabilities into a usable product; it does not include training the underlying models.

How it fits together

The system behind the screen.

  1. 01

    Listing & choices

    A property link, selected photos and narration or format preferences.

  2. 02

    Application service

    A React interface and Node service coordinate the request and record progress.

  3. 03

    Media processing

    A Python service handles background video processing and media assembly.

  4. 04

    Review & download

    Status polling leads to a completed video in permanent storage.

The inspected archive uses React/Vite, Express, a Python media service, Postgres and Supabase. ElevenLabs integration is present in the source; Veo 3 is part of the product workflow described in my résumé.

Engineering & product judgment

Decisions that shape the product.

Put a simple experience over a multi-step job.

The user starts with a listing rather than an editing timeline. Photo choices, narration and output orientation are represented in the generation request.

The tradeoff

Simplicity on the surface creates coordination work underneath: property retrieval, audio, media processing and delivery need a consistent job state.

Let video processing finish in the background.

The application coordinates with a Python service and checks stored progress while processing runs. The completion state exposes the finished video and download.

The tradeoff

Polling is straightforward to operate, but it creates a delay between backend progress and what the user sees. Errors and timeouts need understandable states.

Give progress one shared definition.

The current processing UI reads shared stage definitions and a consistent progress field, including upload and finalization.

The tradeoff

Progress communicates where the job is. It should not pretend to be an exact time estimate when the stages have different workloads.

An iteration from the archive

Fix the progress display, not just the render.

The project’s fix log records two UI components with different step names, inconsistent polling intervals and stages that did not match the backend’s work.

Shared processing constants brought the UI into one sequence. The current modal checks for completion or failure and includes a final upload-and-finish stage.

A video can render successfully while the product still feels broken. Status, recovery and the handoff to the finished asset are part of the experience I’m responsible for.

Outcomes, with context

$10K

Reported MRR

Growth from $0 reported in my September 2026 résumé. This is a historical recurring-revenue milestone, not a claim about today’s revenue.

Solo

Product ownership

Application, integrations and launch. External services supply the underlying infrastructure and model capabilities.

5

Shared progress stages

Verified in the processing UI and its documented correction; not a speed or reliability benchmark.

Sources & scope

Ownership and revenue milestone: September 2026 résumé. Architecture and progress correction: the original Showplace source and implementation notes. No new customer conversion, generation-cost or latency measurement is implied.

Keep exploring

Claude Code Club

Have something in mind?

Say hello.