Aetheria

Aetheria is an AI-powered brand content generator that creates professional social media graphics and captions tailored to your brand identity.

Comment

Inspiration

Every day, millions of small business owners face the same challenge: they need fresh, professional social media content to stay relevant, but they can't afford dedicated designers or expensive agency retainers. I watched local businesses struggle with inconsistent branding, poorly formatted graphics, and the exhausting cycle of content creation that takes hours away from running their actual business.

Aetheria was born from a simple question: What if any business owner could get agency-quality branded content in few minutes?

The vision is democratizing professional design, making it possible for a local bakery to have the same visual presence as a Fortune 500 company.

What it does

Aetheria is an AI-powered brand content generator that creates professional social media graphics and captions tailored to your brand identity.

The workflow is elegantly simple:

Brand Setup — Input your company name, brand color, logo, and optionally a product image

Smart Refinement — AI-generated questions help capture your brand voice, target audience, and content goals

Intelligent Generation — A multi-step AI agent builds a creative brief, designs the visual concept, generates the image, and integrates your logo

Multi-Platform Export — Generate variants in different aspect ratios (1:1 for LinkedIn, 4:5 for Instagram, 9:16 for TikTok/Stories, etc.)

Key Features:

Brand-aware design — Colors, typography, and style match your brand identity

Logo integration — Automatically places your logo appropriately on generated graphics

Product photography — Can incorporate actual product images into marketing materials

Multi-format output — One design, multiple platform-optimized ratios

Caption generation — Platform-appropriate copy with relevant hashtags

Content type flexibility — Social posts or advertisement graphics with CTA overlays

How we built it

Technical Stack: Frontend: React + TypeScript + Tailwind CSS + Framer Motion

AI Orchestration: Gemini 3 Flash with function calling (agentic workflow)

Image Generation: Gemini 3 pro image API

The core innovation is using Gemini as an autonomous creative director. Rather than a simple prompt-to-image pipeline, Aetheria employs a multi-tool agent that:

const agentTools = [ { name: "analyze_brand_identity" }, // Understands brand from inputs { name: "create_creative_brief" }, // Builds structured design direction { name: "generate_background_image" }, // Creates base visual { name: "add_logo_to_image" }, // Composites brand elements { name: "finalize_content" } // Outputs final package ];

The agent reasons through each step, making decisions about:

Visual style (minimalist, bold, elegant, playful)

Color harmony with brand colors

Logo placement (corner, centered, watermark)

Typography style for any text overlays

Composition that leaves space for captions on social platforms

Dynamic Question Generation

Before generation, Gemini analyzes the brand inputs this ensures the AI understands whether you're B2B or B2C, your tone of voice, and specific campaign goals.

Challenges we ran into

  1. Target Audience Ambiguity Problem: The agent couldn't determine if a brand was B2B or B2C, leading to mismatched content tone.

Solution: Implemented dynamic question generation that specifically probes for audience type, industry context, and communication style before generation begins.

  1. Logo Integration Quality Problem: Logos sometimes appeared awkward—wrong size, poor placement, or clashing with the background.

  2. Schema Validation Errors Problem: AI responses occasionally didn't match expected JSON structure, causing runtime errors.

  3. Model Selection Paradox Problem: Initially assumed Gemini 3 Pro would produce better results than 3 Flash.

Discovery: Flash actually generated superior design directions! Pro's verbosity created overly complex image prompts that confused image model, resulting in cluttered outputs. Flash's concise, focused prompts led to cleaner, more professional designs.

Accomplishments that we're proud of

End-to-end generation in mins — From brand input to finished graphic

Truly brand-aware outputs — Not just slapping a logo on stock images

Seamless aspect ratio variants — One generation, infinite formats

Autonomous creative agent — AI makes design decisions, not just follows instructions

Production-ready quality — Outputs businesses can actually post without embarrassment

What we learned

Model size ≠ better results — Gemini Flash outperformed Pro for creative direction. Conciseness in prompts leads to better image generation.

Agentic workflows unlock creativity — Letting the AI orchestrate multiple tools produces more coherent results than chaining isolated API calls.

Context is everything — The questionnaire dramatically improved output relevance. Asking the right questions matters more than having the most powerful model.

Multimodal is the future — Combining text understanding, image generation, and image analysis (for logo placement) in one flow creates experiences impossible with single-modality AI.

What's next for Aetheria Content

Planned Features:

Daily Content Automation — Brands set up their profile once, receive fresh graphics every morning via email

Content Calendar Integration — Connect to Buffer, Hootsuite, or native platform schedulers

Brand Asset Library — Save multiple logos, product images, and brand guidelines

A/B Variant Generation — Generate multiple creative directions to test

Analytics Feedback Loop — Learn from engagement metrics to improve future generations

The vision is transforming Aetheria from a generation tool into a complete AI creative department that handles the entire content lifecycle, from ideation to posting to optimization.

Built With

Share this project:

Updates