Shot Supervisor

A professional-grade, mobile-first AI production companion for planning, generating, editing, and exporting shots with deterministic, studio-level control.

Shot Supervisor bridges the gap between creative intent and technical execution by combining AI planning, JSON-native image generation, and professional post-production tools—directly on a phone.

---

Inspiration

Shot Supervisor was inspired by a real gap in professional visual production workflows: the disconnect between creative intent and technical execution.

Creators today jump between prompts, tools, exports, and manual tweaks just to keep shots consistent across a project. Prompt-based generation is powerful, but unpredictable—and it quickly breaks down when precision, repeatability, or professional standards like HDR, bit depth, and color space matter.

Bria FIBO’s JSON-native, deterministic control felt like the missing piece. Instead of fighting prompts, Shot Supervisor treats AI like a camera department—planned, controlled, repeatable.

The goal: a director-style mobile tool where shots are planned, generated, edited, and exported with production-grade fidelity—anywhere.

---

What It Does

Shot Supervisor is a mobile production workflow app designed for professional creative use.

The app is organized into three core tabs:

Projects

• Create projects with a name and description
• Optionally auto-generate up to 5 AI shots using an LLM-driven breakdown
• Ensures stylistic and technical consistency across a project

Shots

• Each project contains multiple shots
• Create shots with title, description, and generation prompt
• Generate images using Bria FIBO
• Attach reference images from the camera roll or capture photos in-app
• Regenerate shots while preserving intent and structure

Editor

• Professional-grade editing tools:
  • Background removal
  • Background replacement via prompt
  • Background blur
  • Quality enhancement
• Advanced controls:
  • HDR: exposure, contrast, highlights, shadows (custom endpoint)
  • LUTs: None, Cinema (Filmic), Teal & Orange, Vibrant
  • Color Spaces: sRGB, AdobeRGB, DCI-P3, Rec2020
  • Bit Depth: 8, 10, 12, 16-bit
  • Product editing via Bria Product API
• Export options:
  • Save to device or email delivery
  • Formats: JPEG, PNG, TIFF, EXR

This turns a phone into a portable shot planning and post-production workstation, not just an image generator.

---

Bria FIBO Integration (Judges)

All calls to Bria APIs go through custom server endpoints first for postprocessing of format for application.

Example: LLM-generated shot parameters for Bria

{
  "prompt": "Studio-lit product shot of a glass bottle",
  "camera": {
    "focal_length": 85,
    "aperture": 2.8
  },
  "lighting": {
    "type": "softbox",
    "angle": "45deg"
  },
  "color": {
    "space": "DCI-P3",
    "bit_depth": 16,
    "hdr": true
  }
}

Example: Background removal API response

{
  "result": {
    "image_url": "https://d1ei2xrl63k822.cloudfront.net/api/res/88c64c12dfdb4fe9ac0e35e3eea29ed2.png?Expires=1766444665&Signature=..."
  },
  "request_id": "88c64c12dfdb4fe9ac0e35e3eea29ed2"
}

Example: Bria Product API response

{
  "result_url": "https://d1ei2xrl63k822.cloudfront.net/api/res/8a3710a5-b5a5-4467-8ac2-46158a53b171.png?Expires=1766444886&Signature=..."
}

These examples demonstrate deterministic JSON-native control, pre-processing via custom server, and real outputs from Bria APIs.

API integration code is located in:

• /shot-supervisor-endpoints/bria-generate.ts
• /shot-supervisor-endpoints/bria-edit.ts
• /shot-supervisor-endpoints/bria-product.ts

---

How It’s Built

Core Technologies

• Expo / React Native (mobile-first UX)
• Bria FIBO APIs
  • Image generation
  • Editing APIs
  • Product APIs
• Supabase
  • Authentication
  • Projects & shots database

Backend Architecture

Custom Vercel-hosted server endpoints (for security, flexibility, and extensibility):

• bria-generate
• bria-edit
• bria-product
• export-hdr
• hdr-tone-map
• email
• llm-breakdown (uses DeepSeek-V3 to convert project descriptions into structured shot plans with bria)

This modular approach keeps generation, editing, HDR processing, and orchestration scalable and production-ready.

---

Challenges

• Designing a professional editing workflow on mobile without overwhelming users
• Balancing creative freedom with structured, deterministic control
• Integrating HDR and high bit-depth workflows in an intuitive way
• Building custom HDR endpoints beyond default API constraints
• Hackathon time limits—some components are present but not fully polished

---

Accomplishments

• Built a true professional-grade mobile tool, not a demo
• Showcased Bria FIBO’s strengths:
  • JSON-native control
  • Deterministic generation
  • High bit depth & HDR workflows
• Combined LLM-based planning with structured image generation
• Created a scalable, secure backend architecture
• Delivered a full pipeline: concept → generation → edit → export

---

What I Learned

• Professionals don’t want better prompts—they want better controls
• Structured generation unlocks workflows prompt-only models can’t support
• HDR, color space, and bit depth are critical in real production
• Mobile can be a serious creative platform
• JSON-native generation pairs extremely well with agentic pipelines

---

What’s Next

Planned improvements:

• Real-time HDR preview
• Shot consistency auditing
• Histogram-based editing
• MCP-powered Bria conversational edits
• Video support
• Refactor env handling (move all secrets server-side)
• Particle-based splash screen polish

Some features are visible in the codebase but unfinished due to hackathon constraints.

---

Running the Project Locally (Judges)

Prerequisites

• Node.js (18+ recommended)
• Expo CLI
• Android Studio (for emulator) or a physical Android device

1. Install Dependencies

npm install

2. Environment Variables (Client)

Create a .env file at the root of the project:

EXPO_PUBLIC_SUPABASE_URL=your_supabase_url
EXPO_PUBLIC_SUPABASE_ANON_KEY=your_supabase_anon_key
EXPO_PUBLIC_VERCEL_BASE_URL=https://shot-supervisor.vercel.app

EXPO_PUBLIC_VERCEL_BASE_URL points to the hosted server endpoints used by the app.

3. Run a Development Build (Recommended)

eas build --profile development --platform android

Once installed on an emulator or device:

npx expo start --dev-client -c

4. Server Endpoints (Optional – Self Hosting)

If you want to run the backend locally, navigate to:

/shot-supervisor-endpoints

Create a .env file with:

SUPABASE_URL=your_supabase_url
SUPABASE_ANON_KEY=your_supabase_anon_key
BRIA_API_TOKEN=your_bria_api_token
BRIA_API_BASE=your_bria_api_base
GITHUB_TOKEN=optional_for_repo_features
SUPABASE_SMTP_PASS=optional_for_email

Deploy these endpoints to Vercel or run locally depending on your setup.

Notes for Judges

• Some API keys (notably Bria) may expire due to free trial limits
• If you encounter any setup or runtime issues, please contact me—I’m happy to provide:
  • Temporary API keys
  • A live demo walkthrough
  • Clarification on any part of the architecture

This project is best evaluated as a production workflow system, not just an AI demo.

License

MIT