Episode 5

full
Published on:

2nd Sep 2025

Beyond Bananas: Practical Use Cases for Gemini 2.5 Flash

We're setting the stage with a playful correction, Nano Banana is officially Google’s Gemini 2.5 Flash. And Michelle shares why she couldn’t resist testing it for both fun and real-world business use cases, kicking off an honest exploration of what works (and what doesn’t).

[00:01:30] Playing with Creative Images

Michelle experiments with futuristic cityscapes, rebel outfits, and podcast studios. While the results are fun, she calls out perspective challenges and body-type stereotypes still baked into AI outputs.

[00:03:00] Business Use Case: Branding Mockups

Using her Innovative AI Studio logo, Michelle tests Gemini 2.5 Flash for product mockups. The results—like supplement jars and yoga-themed lifestyle shots—prove surprisingly strong. She also notes the hiccups: lag time, token consumption, and blending limitations.

[00:05:30] The Skirt Test: Product Placement Accuracy

Michelle puts Gemini to the ultimate test: a highly detailed parrot-and-palm tree skirt for eBay reselling. The verdict? “Holy shit accurate.” Every detail came through—making this a powerful tool for side hustlers and entrepreneurs needing standout product images.

[00:07:30] Brand Collage & Outfit Creation

Michelle tries building a whole Innovative AI Studio look—leather jacket, distressed jeans, branded cap. While it nails character consistency, Gemini struggles with details like shoe color. Still, the potential for branded visuals is huge.

[00:08:30] Restoring a Family Photo

Michelle uses Gemini to colorize a black-and-white photo of her grandfather. The results? Authentic, realistic, and a meaningful personal win.

[00:10:00] Data Visualization Trials

Michelle explores data viz with PDFs and spreadsheets but finds Gemini’s limitations clear—image cutoffs and token-heavy processing. She pivots to highlight stronger use cases in annotated images and interior design mockups.

[00:12:00] Character Consistency & Upscaling Needs

Michelle highlights how Gemini excels at keeping a model consistent across branded looks but stresses that upscaling is still essential for polished, professional outputs.


Michelle wraps with excitement about Gemini 2.5 Flash’s potential for mockups, product placement, and branding visuals. Her advice: dive in, play with annotated images, and test use cases tied to your business.

Listen for free

Show artwork for Empowered by AI

About the Podcast

Empowered by AI
Women Entrepreneurs Leading the Way
Empowered by AI is the go-to podcast for women entrepreneurs and leaders who are ready to harness AI, no tech background required. Hosted by Michelle Muncy-Silva, founder of Innovative AI Studio and Executive Director of She Leads AI Academy, this show spotlights trailblazing women who are not just using AI; they're demanding a seat at the table in shaping it. From practical tools to ethical frameworks, every episode delivers real talk and actionable strategies for women building businesses and leading boldly in the age of AI. Whether you're automating your workflows, building a standout brand, or navigating the intersection of AI and equity, this is your community. Each episode features conversations with women at the forefront of AI and entrepreneurship, plus the tools, strategies, and stories you can put into practice immediately. We explore AI-enhanced storytelling, using AI to streamline and scale, and what it looks like when women lead the AI revolution on their own terms. Every episode ends with one question to our guests: What's next? Empowered by AI — your weekly dose of inspiration, strategy, and community. Hit subscribe and let's build the future together.

About your host

Profile picture for Michelle Muncy-Silva

Michelle Muncy-Silva

Michelle Muncy-Silva is an AI educator, innovator, and Executive Director of the She Leads AI Academy. She is also the founder of Data Parity AI and host of the Empowered by AI podcast. Her work helps educators, founders and institutions use AI with clarity, ethics, and purpose. With a background in curriculum design and human-centered data practices, Michelle makes emerging technologies practical and accessible for non-technical leaders.