Full-Stack / Mobile / 3D / 2026
Application engineering
From the interface to the data behind it.
I built SceneForge, a presentation builder, and a Flutter records application, connecting interfaces to streamed output and persistent data.
Private applications
{ scene: "pavilion", format: "typed objects" }Follow the data into the interface.
Explore the application paths behind streamed scenes, editable presentations, and mobile reporting. Click a step, connection, or trace to inspect it.
Start the walkthrough, or inspect any step before running it.
Simplified from the project architecture. Runs locally; no live models, credentials, or external services.
The engineering
Inside
the decisions.
Connect the interface to the behaviour underneath it.
Different products need different interfaces, but all need a reliable connection between user intent, application state, backend services, and the final result.
01The design decision
+
Choose the interaction model around the product: records and reporting in Flutter, streamed presentation generation in Next.js, and typed scenes for Three.js.
02The trade-off
+
Streaming changes what the interface must handle. Partial data needs an explicit state, and a scene description needs structure before it becomes something the renderer can use.
03Records and reporting in Flutter
+
Built a Flutter faculty-records application with Supabase integration, records and insights screens, PDF tooling, and automated application builds.
04Streaming a presentation into an editor
+
Developed a Next.js presentation builder with research, outline, card and image-generation routes, streamed generation, and editing interfaces.
05Parsing a scene while it arrives
+
Built SceneForge with typed scene descriptions, incremental JSON parsing, and streamed updates to an interactive 3D scene.
↳Read the system flow
+
Application engineering · Public explanation of the implementation boundaries.
- User intentDescribe a scene, build a presentation, or manage a record.
- Application stateHandle streamed generation, editing, and persistent data.
- Structured outputValidate scene objects, presentation content, or reporting data.
- Usable resultRender editable geometry, presentation cards, or generated reports.
↳SceneForge — streamed 3D editor
+
For someone creating and editing a scene: I built typed scene descriptions, incremental JSON recovery, and streamed updates to Three.js geometry. The live portfolio editor demonstrates object selection, editing, and validated JSON using sample scenes.
↳Presentation builder — generation into editing
+
For someone turning a topic into a presentation: I developed a Next.js application with research, outline, card, and image-generation routes. Streamed output becomes editable application state instead of an opaque final response.
↳Faculty records — mobile reporting
+
For faculty managing records and reports: I built a Flutter application with Supabase integration, records and insights screens, PDF tooling, and automated builds. The challenge connects stored records to readable reports and a usable mobile interface.
↳Project context
+
These are separate private application implementations. Client identities, source, and operational data are not public. No production adoption or deployment scale is claimed.
Private applications
Inspect
the work.
Separate private applications spanning Flutter, Next.js, and Three.js. The engineering connects user interfaces to persistence, streamed generation, structured data, and repeatable builds.
Next project: Polli ↗Source material
Discuss the application work