About this project

This app is built for the FIBO Hackathon to showcase how JSON-native visual generation can give creative teams deterministic control over their images. Instead of wrestling with opaque text prompts, you describe a campaign goal in natural language and the app translates it into structured shot recipes: camera angle, FOV, focal length, lighting, palette, and seeds.

The goal is to make professional, repeatable visual workflows feel simple: art directors can tweak parameters like they would in a DCC or compositing tool, while the underlying Bria FIBO model guarantees consistent, controllable results.

What we built in this project

Our project focuses on making FIBO's JSON-native power usable for non-technical creative teams. Key features we implemented:

  • Goal → JSON shot planner – a guided panel where you describe your campaign (goal, brand colors, tone, region, deliverables) and get back a deterministic list of shot recipes with rationale.
  • Editable pro parameters – for each shot you can fine-tune camera angle, FOV, focal length, lighting description, palette, and seed, with live JSON representation so it is easy to debug and plug into pipelines.
  • Render orchestration layer – a simple bridge that can call a FIBO render API (or fall back to placeholders) so teams can quickly validate recipes and iterate.
  • History scaffold – a dedicated page reserved for storing and revisiting past plans/renders, so in a production version teams could reuse successful setups.

Team

This project was designed and built by a small, focused team combining product thinking, engineering, and visual storytelling:

  • Harshita JoshiLinkedIn
  • Jaivardhan ShuklaLinkedIn