ai3d fab in place

Raindrop services used

• Raindrop Plans & Versioning – Used to define reproducible generation plans for AI-to-3D workflows, enabling consistent iteration and comparison across runs.
  
• Job History & Audit Logs – Provided traceability for asset generation steps, making it easy to debug, validate, and explain results to judges.
  
• Raindrop PRD Generation – Used to auto-generate and refine a lightweight product requirements document directly from the workflow, accelerating iteration under hackathon time constraints.

Vultr services used

• Vultr Cloud Compute – Hosts the backend services and orchestration layer for AI3D Fab-in-Place, chosen for its fast provisioning and predictable performance.
  
• High-Performance NVMe Storage – Supports large 3D assets and intermediate files during AI generation and fabrication prep.
  
• Flexible Pricing / On-Demand Instances – Allowed cost-effective scaling during development without long-term commitments.

Other external services

• Open-source 3D tooling (Blender CLI) – Used for mesh processing, decimation, and baking because it provides industry-standard, scriptable 3D pipelines.
  
• WebGL / Three.js – Used for in-browser preview and placement of generated 3D assets, enabling immediate visual feedback without native installs.
  
• GitHub – Used for source control and open-source collaboration, ensuring transparency and reproducibility.

Why this stack:
Raindrop provided the planning, versioning, and auditability needed to make AI-generated 3D assets trustworthy and repeatable, while Vultr offered a simple, performant infrastructure layer to run the system affordably. The additional open-source tools were chosen for interoperability, speed, and strong ecosystem support.

Built With

• liquidmetal
• raindrops
• vultr