Chaining AI: Planning the Perfect Weekend with Google Labs’ Opal

An 8.5/10 Product Evaluation of the New Visual Itinerary App

Executive Summary:

I stress-tested Google Labs’ Opal to transform a rigid Google Sheets itinerary into a high-hype, photorealistic travel app. By chaining Gemini and Opal, I bypassed data-entry friction to create a premium “concierge” experience for visiting friends. The Result: A stunning 8.5/10 output, despite some experimental hurdles in sharing and consistency.

Context

There is nothing quite like the pressure of having friends visit from out of town. You want to curate the perfect experience, but communicating the plan usually ends up being a messy group chat thread or a bare-bones calendar invite.

Following up from my experiments with building the MTA bus assistant and looking into NYC restaurant health data, I wanted to use AI to solve a personal logistics problem again. This time: creating a high-hype, visually stimulating itinerary for an upcoming weekend with friends.

I recently started using Google Lab’s experimental product, Opal, and I wanted to see if it could handle the heavy lifting of turning my logistics into a custom app-like experience with a polished, shareable output.

User Story

“As a host to out-of-town friends, I want a dedicated app-based itinerary that includes flight details, plans, meals, and photorealistic images, with the ability to export a high-quality PDF to send out and get everyone excited before they even arrive.”

Basic features

• App-Based Interface: A clean, mobile-friendly layout to view the weekend’s plans.
• Logistics Integration: Automatically mapping out specific flight details and restaurant reservations.
• Generative Imagery: High-quality, photorealistic images generated by AI to illustrate the specific “vibe” of our planned stops.
• Shareable PDF Export: A visually rich, stylized PDF that can be dropped into a group chat to build anticipation.

Future Roadmap (V2)

• Weather-Based Recommendations: I plan to eventually integrate real-time weather data to suggest specific packing lists and “what to wear” advice based on the forecast for that specific weekend.

Development method & Pipeline

Instead of feeding raw, messy data directly into Opal, I used a multi-step pipeline to ensure the “app” had the best possible inputs:

1. The Foundation: I started by drafting the core itinerary and flight logistics in Google Sheets (~20 min).
2. The AI Bridge: Rather than manually re-typing everything, I took a screenshot of my Sheets table and dropped it into Gemini to convert it into a clean, structured text itinerary (~10 min).
3. The Final Build: I took that cleaned-up text and used it as the primary input for my Opal app prompt (~15 min).

The Results

Overall, the experiment worked very well. Opal output a beautiful, organized, and fuss-free itinerary. It took my logistical inputs and transformed them into a visual narrative that made the upcoming trip feel real. View the output here

What worked

The core “product” was excellent. The layout was clean, and the photorealistic images added a level of polish that kept the plans clear and the “hype” high.

The Friction points

• No direct sharing URL: You can’t simply send a link to the finished itinerary. Sharing the app link would force friends to generate their own version.
• Export Limitations: Porting to Google Docs and Slides was essentially pointless—the images didn’t port over correctly. I had to use a manual HTML-to-PDF workaround.
• The “Consistency” Trade-off: A notable limitation is the lack of deterministic control. Even using the exact same prompts, Opal’s output varies every time—changing the styling, the layout, and even which itinerary items it chooses to “feature” or focus on.

Product Insight: Personal vs. Scale

While the lack of layout consistency didn’t take away from my specific goal—creating a one-off “wow” moment for my friends—it highlights a major hurdle for broader applications. If you were looking for a large-scale rollout where every itinerary needs a standardized format or “brand” look, current generative tools like Opal might struggle without more rigid template controls.

here’s an alternative visual that was created:

Time it took

Despite the multi-tool workflow, the entire process was efficient:

• Total Time: ~45–55 minutes.

Overall thoughts

I would give this experiment a high 8.5 out of 10. For personal use, the high-quality visual output is worth the manual workaround. It’s a great example of using a “chain” of AI tools to turn a spreadsheet into a premium experience.

TLDR

I used a combination of Google Sheets, Gemini, and Google Labs’ Opal to turn a flight schedule into a stunning app-based itinerary. It’s not quite ready for enterprise-scale consistency yet, but for individual “concierge” hosting, it’s a game-changer.