SNIPR

SNIPR is an AI-assisted renovation workflow platform built by a contractor to manage housing corporation projects from resident intake to floorplans, choices, planning, and execution.

Comment

Inspiration

I own a contracting company in the Netherlands that works extensively with mid-sized housing corporations. In our day-to-day work, renovation projects are not just about planning labor and materials. They involve residents, housing corporation contacts, measurements, room choices, photos, floorplans, communication, execution planning, and documentation.

SNIPR was born from that reality. I wanted a tool that fits how renovation work actually happens in the field: messy, visual, resident-facing, and full of handoffs.

A major goal was to make property inspections as simple as possible. Instead of manually measuring, drawing, photographing, and entering everything into separate systems, SNIPR aims to turn a single room photo into usable project information. With AI, the platform can analyze the image and automatically generate a digital 3D representation of the room as a starting point for renovation planning.

What it does

SNIPR is an AI-assisted renovation workflow platform for contractors working with housing corporations. It helps manage a project from intake to execution:

  • Contractor and corporation administration
  • Demo resident and address workflows
  • AI-assisted room inspection from a single photo
  • Automatic generation of a digital 3D room model
  • Room-based renovation choices
  • Floorplan and room composition
  • 2D and 3D project visualization
  • BAG / address enrichment
  • Resident communication and planning
  • Photos, notes, and execution status
  • AI-assisted workflow support using OpenAI

The goal is to reduce the amount of manual work during an inspection. A field worker can take a photo of a room, after which SNIPR uses AI to recognize the space and create a first 3D room setup automatically. This digital room can then be reviewed, adjusted, and used for renovation choices, planning, communication, and execution.

How I built it

I built SNIPR as a full-stack web application using Nuxt, Vue, TypeScript, Prisma, PostgreSQL, and OpenAI APIs. During Build Week I used Codex intensively to inspect the codebase, fix production issues, improve the demo flow, prepare a GitHub backup, debug 3D and BAG integrations, and make the project easier to share with judges.

Challenges

The biggest challenge was turning real contractor workflows into software without making it feel like a generic dashboard. The app has to support office planning, field execution, visual floorplan work, resident communication, and corporation reporting at the same time.

Another major challenge was simplifying the inspection process. Generating a useful digital room from a single photo requires the system to interpret perspective, walls, room proportions, openings, and visible objects. The result must remain editable, because real renovation projects always contain exceptions and incomplete information.

Another challenge was integrating real-world Dutch building data such as BAG, address coordinates, and 3D building context. These external services can be slow or inconsistent, so I had to make the app resilient and useful even when some data is delayed.

What I learned

I learned how powerful Codex can be as a practical engineering partner, especially inside an existing, messy, real-world product. It helped me move faster, but also helped me reason through bugs, deployment, demo data, and product positioning.

I also learned that AI becomes most valuable when it removes practical work from an existing process. For SNIPR, that means helping a contractor move from a room inspection to a usable digital project environment with as few manual steps as possible.

What's next

Next I want to further improve the one-photo inspection workflow, make the automatically generated 3D rooms more accurate, and allow users to review and correct the result more easily.

I also want to improve the contractor demo experience, add more robust AI-assisted project summaries, deepen the floorplan and 3D workflow, and make SNIPR easier for other contractors and housing corporations to test.

Demo access

Live demo: https://zoov.snipr.nl

Demo login: Email: demo@snipr.nl Password: snipr-demo-2026

GitHub repository: https://github.com/sakeverpalen/SNIPR

The GitHub repository is private. I invited build-week-event@openai.com and devposttesting as collaborators on GitHub. The invitations are currently pending.

Built With

Share this project:

Updates