NotCo – Brazilian Vegan Product Visualization
What It Is
I'm testing the adaptation of prompts and styles with things that I don't see that often: vegan products that Brazilians can find on the shelves. It took me a few credits to get this one right, but I'm pretty happy with the results. This project explores the challenge of generating AI content for products that are underrepresented in training data—specifically, Brazilian vegan products. The workflow involved using Perplexity for market research and product understanding, ChatGPT for crafting culturally-aware prompts, and Google Flow for generating the final video content. The result is product visualization that feels authentic to the Brazilian market while showcasing vegan alternatives in an appealing, relatable way.
Features
As the AI orchestrator and creative strategist for this product visualization project, my responsibilities included researching the Brazilian vegan product market to understand what's actually available to consumers, developing prompt strategies that could generate authentic representations of less-common product categories, iterating through multiple generations to refine visual quality and market authenticity, coordinating a multi-tool workflow (Perplexity, ChatGPT, Google Flow), and ensuring the final output resonated with Brazilian consumer preferences and shelf realities.
Technical Highlights
Market Research with Perplexity
Leveraged Perplexity's research capabilities to understand the Brazilian vegan product landscape, identifying specific brands, product types, and packaging styles that Brazilian consumers encounter in stores.
Culturally-Aware Prompt Engineering
Used ChatGPT to craft sophisticated prompts that account for Brazilian market context, consumer preferences, and the specific visual language of vegan products available locally—adapting general prompts to specific cultural and commercial realities.
Google Flow Video Generation
Utilized Google Flow's video generation capabilities to create dynamic product visualizations that bring static product concepts to life with movement, lighting, and appealing presentation.
Iterative Refinement
Went through multiple generation iterations to achieve the right balance of authenticity, appeal, and market relevance—demonstrating the importance of refinement in AI content creation for specific markets.
Underrepresented Content Challenge
Tackled the unique challenge of generating quality AI content for product categories that are less common in training data, requiring creative prompt strategies and multiple attempts to achieve satisfactory results.
Multi-Tool Workflow
Demonstrated an efficient workflow combining research (Perplexity), prompt engineering (ChatGPT), and content generation (Google Flow) to produce market-specific marketing content.
Key Features
PROJECT TYPE: Product Visualization Experiment
MARKET FOCUS: Brazilian vegan products
CHALLENGE: Generating content for underrepresented product categories
TOOLS: Perplexity, ChatGPT, Google Flow
WORKFLOW: Research → Prompt Engineering → Video Generation
ITERATIONS: Multiple refinements to achieve market authenticity
CONTEXT: Products actually available on Brazilian store shelves
Results
Successfully created authentic product visualization for vegan products in the Brazilian market context, an underrepresented category in AI-generated content.
Developed a replicable multi-tool workflow for generating market-specific product content: Perplexity for research, ChatGPT for prompt engineering, Google Flow for video generation.
Demonstrated that AI can effectively visualize niche or underrepresented product categories with proper research, prompt adaptation, and iterative refinement.
Proved the importance of cultural and market context in AI content generation—showing that generic prompts often fail for specific local markets.
Created compelling product marketing content that resonates with Brazilian consumer preferences and accurately represents available vegan alternatives.
Established best practices for adapting AI generation tools to local market realities, particularly for products that may not be well-represented in training data.
Showcased the value of persistence and iteration in AI content generation—demonstrating that quality results for specific contexts require multiple attempts and refinement.
Validated that AI-generated product visualization can serve real marketing needs for emerging product categories in specific geographical markets.
Provided a framework for other creators working with underrepresented product categories or specific market contexts in their AI-generated content.