The Journey of Developing Flex Reminder: When a Backend Developer Teams Up with AI

The Beginning: A Personal Problem Turns into a Project

“Let me save this video for later…” “I’ll add this article to my list…” How many times have you said that — and never gone back?

That was my struggle. My “saved” lists kept growing endlessly, while valuable content got buried under good intentions. As a backend developer, I didn’t have the mobile development expertise to build the solution myself — until AI opened a new door.

Flex Reminder was born from that need: an app that gathers all the content I want to revisit and uses AI to determine the best time to remind me about it — smartly and repeatedly — until I actually open it. No more forgotten lists.

The Strategy: Dividing the Work Between AI Platforms

Being solely a backend developer posed a major challenge: How could I build a professional Flutter app with no prior experience?

The solution was to strategically leverage the strengths of different AI platforms:

1. Claude: The Thinker and Architect

Why Claude? From my experience, Claude excels at complex reasoning in programming. Its ability to understand architectural problems and detect logical issues is unmatched.

How I Used It:

• Architectural Planning: Designing the app’s structure and choosing the right state management system.
• Feature Selection: Deciding which features were essential and which could be postponed.
• Debugging: When something broke, Claude would analyze the code and pinpoint the cause precisely.
• Logic Review: Ensuring the business logic was efficient and consistent.

The Challenge with Claude: Claude tends to produce concise, summarized responses — great for planning and reasoning, but not ideal for writing full code files. I needed another partner for hands-on implementation.

2. Gemini: The Executor and Code Engine

Why Gemini? Here’s the clever part: Flutter is built by Google — and so is Gemini. That means Gemini has deeper, more up-to-date knowledge of Flutter and its packages.

How I Used It: Once I planned with Claude, I transferred the requirements and architecture decisions to Gemini:

• Full Code Generation: Gemini wrote complete Dart files with detailed structure.
• Flutter Widgets Expertise: It had excellent understanding of the latest widgets and Flutter best practices.
• Package Integration: Gemini knew exactly how to implement Provider, Riverpod, and other common packages.

Personal Experience: The transition between Claude for planning and Gemini for execution was surprisingly smooth — like having a senior architect (Claude) and a skilled developer (Gemini) on the same team.

3. FlutLab: The Beginner-Friendly Development Environment

Why FlutLab? As a Flutter beginner, the last thing I wanted was to spend days setting up Android Studio, SDKs, and emulators.

What FlutLab Offered:

• Instant Environment: No installation, no setup — just open the browser and start coding.
• Web Emulator: I could instantly preview the app in a browser without needing a phone or Android emulator.
• Accessibility: I could work anywhere, from any device, as long as I had internet.

The Achievement: I managed to complete about 70% of the app using FlutLab — building the UI, the basic state management, and the core features.

The Limitations: Once development advanced, I needed features beyond FlutLab’s capabilities:

• Firebase Messaging: To send reminders to users.
• Firebase Auth: For user account management.
• Real Device Testing: Essential for real-world validation.

While perfect for beginners, FlutLab wasn’t built for deep cloud integrations.

4. Firebase AI Studio: The Game Changer

Why Switch? As development progressed, integrating Firebase became essential. Firebase AI Studio (formerly Firebase Extensions, now part of Google Cloud AI) provided the next-level tools I needed.

New Capabilities:

• Direct Firebase Integration: Smooth connection with Cloud Messaging, Auth, and Firestore.
• Broader Testing Environment: Realistic push notification and authentication testing.
• Advanced Development Tools: Better debugging, clearer logs, and production-ready deployment.

Personal Experience: Moving to Firebase AI Studio felt like stepping from the training field into a professional stadium. Suddenly, the app could send real notifications, manage real users, and behave like a true product.

Bonus: Because Firebase AI Studio is part of Google’s ecosystem, integration with Gemini became seamless — I could now use it to write Cloud Functions, configure Firebase Rules, and optimize Firestore queries.

Technical Side: Challenges and Solutions

Challenge 1: State Management

The Problem: As a backend developer, concepts like StatefulWidget and Provider were new territory. The Solution: I used Claude to understand the differences between Provider, Riverpod, and Bloc, then asked Gemini to write a full implementation for the chosen method.

Challenge 2: UI/UX Design

The Problem: I had no background in interface design. The Solution: Claude helped me study best practices for productivity apps, while Gemini translated those ideas into functional Flutter widgets.

Challenge 3: The Smart Reminder Algorithm

The Problem: How can the app determine the optimal time to remind the user? The Solution:

• Claude designed the logic, factoring in user free time, content priority, and deferral frequency.
• Gemini coded the algorithm in Dart.
• Firebase AI Studio executed scheduled functions to trigger reminders at the right times.

Challenge 4: Firebase Integration

The Problem: FlutLab couldn’t properly test Firebase Auth or Messaging. The Solution: Transitioning to Firebase AI Studio allowed me to:

• Configure Firebase with proper credentials and settings.
• Test push notifications on real devices.
• Apply Security Rules to protect user data.

Lessons Learned: Tips for Backend Developers Entering Mobile

1. Choose the Right Platform for Each Stage

• FlutLab → Perfect for prototyping and early-stage UI work.
• Firebase AI Studio or local environment → Essential for advanced integrations.

2. Leverage Each AI’s Strengths

• Claude → For planning, architecture, and debugging.
• Gemini → For implementation and Google-specific technologies.

3. Don’t Fear Switching Tools

Moving from FlutLab to Firebase AI Studio wasn’t failure — it was natural progression. Each tool served its purpose at the right time.

4. Learn the Core Concepts

Even with AI help, understanding the fundamentals (Widgets, State, Lifecycle) is critical to asking the right questions.

5. Test on Real Devices Early

Simulators are useful, but real devices reveal issues you won’t see in web previews.

Conclusion: The Future of AI-Assisted Development

Flex Reminder wouldn’t exist without AI. As a backend developer with no Flutter experience, I was able to build a fully functional app thanks to:

1. Strategic Planning with Claude — for architecture and problem-solving.
2. Fast Implementation with Gemini — for actual Flutter code and Google integration.
3. Easy Start with FlutLab — for beginner-friendly development.
4. Professional Finish with Firebase AI Studio — for full-scale integration and deployment.

This experience proves that entry barriers into new domains are lower than ever. AI isn’t just a tool — it’s a true collaborator that bridges skill gaps and turns ideas into reality.

Final Message: If you have an idea worth building, don’t let lack of expertise hold you back. The tools are available, the knowledge is accessible, and AI is ready to help you — start today.

Note: The app is still under development — the journey continues. Every day brings new lessons and challenges, each unlocking fresh, creative solutions. That’s the joy of building in the age of AI.

https://play.google.com/store/apps/details?id=com.saadmohammed2000.flex_reminder