[Day 2] Nano Banana Pro: Zero to Infographic in 10 Minutes
[Day 2] Nano Banana Pro: Zero to Infographic in 10 Minutes
A step-by-step experiment using Gemini’s Nano Banana Pro for AI image generation. See the prompts I used, how I iterated with design trends, and the workflow that took me from concept to a ready-to-share infographic in one chat session.
What if you could go from a rough idea to a polished, professional infographic in under ten minutes—without touching design software? That's exactly what I tested today using Google Gemini's Nano Banana Pro image generation feature. As someone working at the intersection of HR, industrial-organizational psychology, and technology, I'm constantly looking for ways to communicate complex ideas visually. The results surprised me: not just in speed, but in accuracy—including text that actually reads correctly, which has been a persistent pain point with other AI image generators.
This experiment reinforced something I believe deeply: AI is a powerful amplifier of human creativity, but it still requires a human in the loop—someone who can ask the right questions, apply domain knowledge, and exercise taste in refining outputs. Let me show you exactly how this played out.
Today's Experiment
Goal: Create two AI-generated visual assets using Gemini's Nano Banana Pro—a blog thumbnail and a workplace wellness infographic—while testing iterative refinement and style customization.
Tools Used: Google Gemini (with Nano Banana Pro), Perplexity AI (for design trend research)
Time Investment: Approximately 10 minutes from initial prompt to final outputs
The Process
Step 1: Accessing Nano Banana Pro
In the Gemini app, I selected the banana icon to indicate I wanted to create an image. The interface is straightforward—you simply describe what you want, and Gemini handles the generation.
The Gemini home screen with the 'Create image' option highlighted
Step 2: Crafting the Initial Prompt
My first prompt asked for two outputs: a thumbnail showcasing Nano Banana Pro's capabilities, and an infographic that would be easy for beginners to follow. I intentionally started broad to see how well Gemini would interpret my intent.
My initial prompt requesting both a thumbnail and an infographic
Step 3: Reviewing First Results
Gemini generated both images with impressive speed. The thumbnail showed the concept of text-to-image transformation, and the infographic broke down the process into three clear steps. Importantly, the text in both images was legible and spelled correctly—something that has historically been a major weakness of AI image generation.
The initial outputs: a thumbnail demonstrating text-to-image and a step-by-step infographic
Step 4: Iterating with Context
Here's where human judgment became essential. Rather than accepting generic outputs, I pushed Gemini to incorporate real-world context. For the thumbnail, I asked it to visualize a futuristic Singapore based on the actual Master Plan 2025 (MP2025). For the infographic, I shifted to a workplace application: creating a wellness infographic for HR that would promote our company's #wellness Slack channel.
Refining the prompt with specific, contextual requirements
The results were remarkable. Gemini pulled in themes from the MP2025—biophilic design, green corridors, car-lite zones—and rendered them into a cohesive futuristic cityscape. The infographic now featured HR-specific use cases and behavioral science tips.
Refined outputs incorporating Singapore's Master Plan and HR wellness context
Step 5: Applying Design Taste
This is where curation skills come into play. Using Perplexity AI, I researched current graphic design trends and found two that appealed to me: Digi-Cute (kawaii-inspired with pixel art elements and bright gradients) and Elemental Folk (traditional folk patterns with modern layouts). I tested both styles on the infographic.
The HR guide rendered in Digi-Cute style with kawaii elements
The same content in Elemental Folk style—warmer and more organic
Step 6: Creating the Final Deliverable
I chose the Elemental Folk style for the final wellness infographic, as its warmth aligned better with the 'New Year, New You' message. The final piece included three behavioral science-backed tips: starting with micro-habits, anchoring to identity, and practicing self-compassion. It also featured a clear call-to-action to join the #wellness Slack channel.
The final wellness infographic: 'New Year, New You: Cultivating Wellness Habits'
Step 7: Generating Accompanying Copy
Without leaving the same Gemini chat, I asked for draft Slack messages to accompany the infographic. Gemini provided three options with different tones—energetic, concise, and community-focused—each with appropriate emojis and a call-to-action. This demonstrates the power of keeping everything in a single workflow: idea → visual → copy, all in one conversation.
Multiple message options generated in the same conversation
Outputs
From this single experiment, I produced:
A blog thumbnail featuring a futuristic Singapore cityscape based on the Master Plan 2025, complete with biophilic architecture and green corridors
A workplace wellness infographic with behavioral science tips for building new habits, designed in an Elemental Folk style with a warm, approachable aesthetic
Three ready-to-use Slack messages with different tones, each including the call-to-action for the #wellness channel
A replicable workflow that demonstrates how to go from concept to polished communication materials in 10 minutes
What I Learned Today
Text accuracy has dramatically improved. Unlike my experiences with ChatGPT's DALL-E or Meta AI, Gemini's Nano Banana Pro consistently produced legible, correctly spelled text in the images. This is a significant leap forward for anyone needing infographics or diagrams with labels.
Context and iteration are everything. The first outputs were generic. By adding specific context—Singapore's urban planning documents, HR wellness applications, design trend names—the results became genuinely useful and tailored.
Taste is the human differentiator. AI can generate infinite options, but choosing between Digi-Cute and Elemental Folk required understanding the audience, the message, and the emotional tone I wanted to convey. This is where a curator's eye becomes invaluable.
Single-chat workflows accelerate execution. By keeping the thumbnail, infographic, and copy generation in one conversation, Gemini maintained context throughout. This dramatically shortened the path from idea to execution.
AI research complements AI creation. Using Perplexity AI to research design trends before applying them in Gemini showed how multiple AI tools can work together—one for knowledge gathering, another for creation.
Try It Yourself
Ready to test this workflow? Here's how to get started:
Open Google Gemini and select the banana icon (🍌) to activate image generation mode.
Start with a clear prompt that specifies what you want to create and its purpose. Example: "Create an infographic that explains [topic] for [audience] with clear steps and visual icons."
Iterate with context. Don't accept the first output. Add domain-specific details, reference real documents or plans, and specify the intended use case.
Research and apply design styles. Use Perplexity AI or a similar tool to discover current design trends, then ask Gemini to apply specific styles by name (like 'Digi-Cute' or 'Elemental Folk').
Generate accompanying copy in the same chat. After creating your visual, ask Gemini to write the message, email, or social post that will accompany it. The context carries forward.
Key Prompt Template: "Create a [type of visual] for [audience] that shows [main message]. Use [style name] design. Include [specific elements like text, icons, steps]. The visual should convey [desired emotion/tone]."
This experiment reinforced my core thesis: AI doesn't replace human judgment—it amplifies it. The speed from idea to polished output was remarkable, but every meaningful improvement came from human decisions about context, style, and purpose. That's the future of work I'm excited about: humans and AI collaborating, each contributing what they do best.
Tomorrow, I'll explore another tool and workflow. Follow along as I document this year-long journey of daily AI experiments.
Every year since ChatGPT launched, I've found myself asking a completely different question about generative AI—not the same question with better answers, but an entirely new question that only became possible because the previous one had been resolved. In 2022, I wondered what AI actually knew (and learned the hard way about hallucinated salary data in job descriptions). By 2025, I was producing interactive training decks and Excel templates without writing a single line of code. Now, at the start of 2026, I'm asking how to build automated workflows that free me up to focus on what humans do best. These five questions trace the extraordinary pace of change in just five years—and reveal where the real opportunities lie for knowledge workers ready to ask the next question
I've tried using Claude, ChatGPT, and Gemini to create presentations and the results were always underwhelming. The content was decent, but the formatting? A manual nightmare. So when I tested Gamma.app with a genuinely complex use case (employee offboarding compliance across six Asia-Pacific countries), I wanted to see if a specialized tool could actually deliver what general-purpose AI assistants can't: polished, presentation-ready output in minutes, not hours. Here's exactly what happened, step by step, including the limitations they don't tell you about upfront.
What happens when you give three AI tools the same complex virtual training scheduling problem and one of them walks straight into a cultural landmine? I ran an experiment with Claude, ChatGPT, and Gemini: identical prompts asking each to recommend the best approach for scheduling a 100-person virtual training across six APAC countries during Lunar New Year season. One tool recommended a culturally tone-deaf date to save $2,500, and one framed the time zone constraint as a "Golden Window" that actually helped me think through the problem differently. This is what AI collaboration looks like in practice and why the human still needs to be in the loop.