Ethni-CITY: The New Travel App

Ebuka Achibiri — Sun, 15 Mar 2026 18:22:52 +0000

If you're chronically online like I am, you're used to seeing tons of reels of people who travel to all these "authentic" destinations: Sao paulo, Marrakech, Cairo, Bangkok, and Bali. On Instagram or TikTok, it's pretty common to accompany your posts with a soundtrack, and it's the norm to post the latest trending songs to garner likes and reposts.
No shame here.
We all want validation, but so do the local economies we're coming from, and they hardly benefit when our money goes to international hotel chains and global brands. However, there are a few ways we can support locals while still making sure your pics eat, and one of them is by supporting local artists.
With Ethni-CITY, we connect your pics to the local artists of the countries you visit to curate a niche storyboard that puts artists' songs on your posts, helping them monetize their craft, gain recognition, and in turn boost their micro-economy as your pics garner likes.

To build this, I weaved spatial data with cultural analysis.

Using Gemini 3.1 pro preview as the main model, the app performs a deep multimodal analysis of your uploaded photos and then identifies the city and country the picture was taken in by looking for landmarks, cultural motifs, textile patterns and lighting conditions to determine.
I used Cesium JS along with Google Cloud photorealistic tiles, the agent transports you to the location of the photo, helping to tell a centric story.
To ensure the app survives the "Hackathon Hug," I implemented a hierarchical fallback logic:
Tier 1: gemini-3.1-pro-preview for high-fidelity creative direction.
Tier 2: gemini-2.0-flash for high-speed fallback.
Tier 3: gemini-1.5-flash-8b for emergency quota resilience.

CHALLENGES FIXED:

By integrating a heavyweight library like Cesium into NextJS App router environment, posed significant challenges with asset serving and Typescript definitions. I had to automate the movement of Cesium's build assets into the public directory during the pre-build phase to ensure the tiles are rendered correctly in production.

I'm grateful for Gemini for providing the tools to curate this experience and help promote local artists from around the world.

Explore the project on GitHub:[ https://github.com/david-ac1/Ethni-CITY]
Watch the demo:[ https://youtu.be/kdrUzPtRw9Y?si=EzWt6DvbdbLjOk4V]

#GeminiLiveAgentChallenge #GoogleAI #GoogleCloud #Gemini

Why Most Health-Tech Architectures Fail the Policy Test (and What I'D Do Instead)

Ebuka Achibiri — Sat, 03 Jan 2026 12:50:59 +0000

Most people in tech are obsessed with "scaling." They want to talk about Kubernetes clusters that can handle millions of hits or AI models that "disrupt" diagnostics.

But when you look at it through the lens of Health Economics, that kind of "innovation" is usually just a massive hidden tax.

I’m currently focused on Digital Policy, and the more I look at the "hyperscale" cloud, the more I see a trap. If you’re building a health exchange in a market with strict data residency laws, AWS isn't your friend—it’s a liability. Why? Because complexity is the enemy of an audit. If your infrastructure is so "sophisticated" that a policy-maker can't understand where the data lives, you’ve already lost the trust of the people you're trying to serve.

We need to start advocating for what I call "Sovereign Simplicity." This means:

Regional Pinning over Global Mesh: If the law says data stays in-country, your network architecture should be physically incapable of moving it. That's not a "limitation," it's a feature.

Predictable Billing as a Safety Feature: In health-tech, a "spike" in your cloud bill shouldn't be the reason a clinic goes offline. Flat-rate, predictable infrastructure (like the kind you get with Droplets or simple VPS setups) is a moral choice, not just a financial one.

The goal shouldn't be to build the most "advanced" system. It should be to build the system that is the easiest to explain to a regulator and the cheapest to keep online for a decade.

In the world of health data, boring is better. It’s safer, it’s cheaper, and it’s the only way to actually scale a policy-compliant system in the real world.

Forem: Ebuka Achibiri

Ethni-CITY: The New Travel App

Why Most Health-Tech Architectures Fail the Policy Test (and What I'D Do Instead)