VRon
Contact us
Sign in
September refactored subscriptions and S3, hardened auth, and finished the iOS scanning engine—plus a first look at AI Content Factory for SME reels, images, and later 3D from photos or text.
engineering, platform architecture, subscriptions, s3, ios, lidar, ai, content factory, sme
BlogMonthly Update: Subscription Engine, S3 Storage, iOS Scanning & AI Content Factory

Monthly Update: Subscription Engine, S3 Storage, iOS Scanning & AI Content Factory

Sep 30, 2026By VRon Team
iOS Roomscan details: a stitched interior mesh with furniture excluded, exported as USDZ

September was an architecture month at vron.one. Behind the scenes we overhauled core API services, decoupled media file handling, and stabilized billing so spatial commerce can run at a larger scale. On the product side, the iOS scanning engine is ready—and we are hinting at a future side project for small and medium companies.

Subscription engine refactor (SubscriptionService)

To support dynamic pricing, multi-currency checkout, and custom Bring Your Own World (BYOW) annual plans, we pulled all subscription logic into a dedicated, unit-tested SubscriptionService (#47):

  • Robust Stripe webhook pipelines and metadata sync (#90, #91) so renewals and schedules stay accurate and event loops do not fire twice.
  • Multi-currency support (#90) across global merchant tiers.
  • Self-service downgrades that respect BYOW world limits when you step down a plan.

Decoupled S3 storage and product services

As spatial assets grow, file ownership and paths need clean boundaries:

  • A central fileService (#85) now owns S3, path prefixes, task scheduling, and user URI retrieval.
  • Product services handle background spatial generation, thumbnail creation, and cascading media hard-deletion (#82) without leaking storage rules into every feature.

Auth security and session hardening

  • Hardened authentication cookies (#88).
  • Strict cookie clearing on sign-out across subdomains, so landing, merchants, and related apps do not overwrite each other.
  • Reinforced UserAuthGuard on backend API endpoints.
  • The public landing site no longer forces a login before you can read the blog or browse worlds.

App spotlight: iOS spatial scanning

Core engine work on the iOS Spatial Scanning App is complete. Using Apple LiDAR and native spatial frameworks, you can scan full interiors in minutes, generate multi-room stitched GLB (and USDZ) models, and upload them straight into vron.one.

The Roomscan details screen (hero image) shows a typical output: a clean structural mesh, scan mode Standard, furniture not included, and a derived “excluded-furniture” source when you ask the pipeline to strip clutter.

Coming later: AI Content Factory

We are planning a side project for small and medium companies: AI Content Factory.

The first job is everyday marketing, not a new 3D suite. SMEs will be able to produce their own reels, short videos, and images without a full studio. Think product shots, property teasers, and social cuts you can post the same week.

Later, the same factory is meant to feed vron.one:

  • 3D objects from photos
  • 3D from text (describe the piece)
  • 3D from text plus photos when you want both a prompt and references

That pipeline is not live. It will grow beside the core platform. Stay tuned.

Lisbon is next

Web Summit 2026 in Lisbon is the next public stop. We will show the scanning path—physical space to interactive GLB—and the quieter architecture that keeps subscriptions and storage honest.

See you there—or on the blog in October.