JENI — Promptless Studio

💡 Inspiration

Modern AI image generation relies heavily on prompt engineering — an approach that is often unpredictable, inconsistent, and difficult to scale for professional workflows. Small changes in wording can drastically alter outputs, making it unsuitable for production environments where consistency, repeatability, and control matter.

While exploring Bria FIBO, we were inspired by its JSON-native, deterministic control system, which rethinks how creators interact with generative models. Instead of guessing the “right prompt,” creators can define explicit parameters like camera angle, field of view, lighting style, color palette, and output quality.

✨ JENI — Promptless Studio was born from the idea that UI-driven, structured control should replace prompt engineering, especially for professional creative workflows.

🧠 What it does

JENI — JSON Engine for Imagery is a promptless visual creation studio that allows creators to generate AI images using structured JSON controls instead of text prompts.

Creators interact with a clean UI to:

🎥 Choose camera angle and field of view
💡 Define lighting style and rim light
🎨 Lock color palette and visual mood
🖼️ Select output resolution, HDR, and bit depth

Every interaction updates a live JSON schema, which is sent directly to the Bria FIBO model to generate deterministic, production-ready images.

🚫 There is no free-text prompting — visuals are driven entirely by structured parameters, ensuring repeatability and professional-grade control.

🛠️ How we built it

Designed a prompt-free UI using form controls, sliders, toggles, and dropdowns
Mapped every UI interaction directly to a strict JSON schema
Integrated Bria FIBO’s text-to-image API, using structured parameters instead of prompts
Implemented a live JSON preview panel to make the generation logic transparent
Built the experience as a frontend-first prototype focused on UX clarity

💙 Lovable was used to rapidly prototype and iterate on the interface, enabling fast experimentation with layouts, controls, and interaction flows.

⚠️ Challenges we ran into

Translating creative intent into structured controls without losing expressiveness
Designing a UI that feels intuitive while exposing professional-grade parameters
Ensuring the JSON schema stayed clean, deterministic, and extensible
Balancing simplicity for new users with depth for advanced workflows

One key challenge was resisting the temptation to “fall back” to prompts — staying fully promptless required deliberate design decisions.

🏆 Accomplishments that we’re proud of

Built a fully promptless image generation workflow
Demonstrated FIBO’s JSON-native strength in a clear, visual way
Created a system where outputs are repeatable and controllable
Delivered a professional, production-oriented UX, not just a demo toy
Clearly differentiated from prompt-based tools in both concept and execution

📚 What we learned

Structured control unlocks trust and reliability in generative AI
UI design is just as important as model capability in creative tools
Deterministic systems are far better suited for enterprise and production use
FIBO’s architecture enables workflows impossible with prompt-only models

✨ Most importantly, we learned that removing prompts doesn’t reduce creativity — it refines it.