Backstory: QRMenuConnect.com My Journey: How I got here to launch an app by Bootstrapping.
“From IT civil service to startup, my path was bumpy to say the least, but every coding scramble was a step closer to helping independent restaurants bring their menus into the digital age.”
After an unexpected end to my career as Director of Information Technology for a New York City government agency in early May 2024, a devastating blow to me and my family; I was determined to start a new chapter. One where I could build tools I either needed myself or knew others had a need as well. Without any resources or an app developer background I hunkered down. My life was and is filled with challenges, so what is one more thing I needed to overcome. One of those projects became QR Menu Connect, a web app designed to support the restaurant industry, one of the sectors hit hardest by rapid digital transformation. While not originally the focus of my small consulting tech business I planned to start while sending out resumes to correct the fallout from a termination (I should explain why and how regarding but, this is not the time), this project became a key addition, driven by both necessity and curiosity to build a better tool.
With very limited, or I really should mention no funds at all, but an intense desire to make something, I dove into the world of AI-powered coding assistant tools. I began as a true bootstrapper, using web-based IDEs like Bolt, Lovable, Cursor, Tempo Labs, and Replit. Eventually, I transitioned to more advanced setups, coding in local IDEs like VS Code and integrating APIs with tools like Anthropic, Cline, and Gemini. A goal of mine was to deploy my apps to Docker containers hosted on my servers, but I put those plans aside due to the complexity of trying to get a fully functioning application launched. My thought process as it always is to save cost when you can build it yourself and host yourself.
By early 2025, my very small budget, roughly $250 (as I said it was very little, hey that’s bootstrapping), began to spiral out of control, largely due to heavy experimentation and hallucination with AI LLMs, particularly Anthropic’s API, which—while powerful at the time, quickly became unsustainable on a limited budget. The costs were simply too high for an independent developer without outside funding, not to mention the loss of income putting pressure on my family’s finances. I had to make the difficult decision to stop using Anthropic API and deployment plans, so I had to find a free solution. That’s when I decided to take another look at Google’s Firebase.
As the months went by since my last testing session, Gemini Flash had improved, and I read that Gemini 2.5 Pro had gotten significantly better.
I don’t know the exact number of versions the app went through—let’s say about eight—before I returned to Firebase Studio and Gemini Flash. As I started with Firebase Studio, I noticed that Gemini isn’t very creative on the UI side, especially with color and visual appeal—but it was getting the job done. I don’t want to get too technical, but there was a major setback when I had to scrap a functioning test environment version due to (believe it or not) a Google Cloud Service integration conflict with Google Authentication logins. The code just got too messy and had to go.
After starting the project from scratch again, another major roadblock was deployment. A deeper dive into a technical tale for another day—simply what happened was the Firebase Studio app, when attempting to deploy, the configuration was referring to the old backend Firebase project. During the chaos of trying to resolve this issue, I had to switch to Gemini 2.5 Pro, as the code had simply become too sophisticated for Flash. I told myself, “I have no funds to really do this, but I am so close to launching this app.”
It took time, but I learned a lot in the process—how Firebase Studio works with app development, Google Cloud services, APIs, secret management, GitHub pushes, and deployment to app hosting all within the Firebase environment. In the end, using Google’s Firebase provided a pay-as-you-go database that helps when you have several coding app projects in a dev environment and no need for a monthly fee to have multiple backend databases active. That is tremendously helpful when you have different versions and testing ideas of other app projects. I’m jumping ahead of myself, but QR Menu Connect is not the only real-world project to launch.
Although I have years of experience hosting back-end infrastructure—including running hosting services and managing multiple websites on my own Linux servers at colocation facilities—I didn’t choose Firebase for its hosting capabilities. I chose it specifically for the Flash AI code-assist freebie and pay-as-you-go database hosting. Since I was bootstrapping this project, that went a long way. It gave me a valuable edge during rapid development phases. Despite failures and my budget, Gemini 2.5 Pro pulled through with complex code logic and deployment to Firebase App Hosting. The only thing in the back of my mind is migration to another backend. Using Firebase really locks you in—but Firebase did allow the idea to come to life, so I am loyal to the Fire.
As mentioned, the app grew in size and complexity. With multi-level user permissions, subscription tiers, 8.5×11 and 4×6 flyer printing capabilities, and advanced API integrations, the Flash version of Gemini (Firebase’s default AI assistant) simply couldn’t keep up. I eventually upgraded to Gemini Pro 2.5 in July, which provided deeper context understanding and more reliable support for managing the app’s increasing sophistication. That shift was crucial in helping me continue building without burning time or resources.
I need to put it out there: all tutorials online—especially those on YouTube—focus on simplified apps. They rarely prepare you for what it’s like to build a real functioning, full-stack application. QRMenuConnect required:
- User permission management (admin, restaurant owners)
- Subscription-level features (Basic Free vs. Pro Paid access)
- Integration with third-party APIs, such as:
- Stripe for payments
- SendGrid for transactional emails
- Google Cloud Services for user verification, database, and storage
- Document generation features like printable QR code flyers
- and Text-to-speech for the visually impaired
These features required a more mature development approach and a well-structured backend, not something you often find in “build an app in 20 minutes and make $10,000 a month” YouTubers crank out by the dozen.
In April 2025, I leaned further into Firebase Studio (I had no choice with no extra funds)—not as a hosting solution as I stated before, but as a smart AI-driven development companion. Firebase also included database and storage with pay-as-you-go pricing, which helped. Every tool became essential not just for cost savings, but for staying nimble through the many iterations of my app. At times, AI-generated code forced me to rebuild from scratch. Through long nights of debugging delirious code logic gone wild, I discovered the quirks of coding with AI as both a partner and an unpredictable co-pilot on a bender trip. In the end, Firebase did make it happen, and I was not so positive with the solution. In fact, I may have written it off as bloated, dumb, and a waste of time.
The heart of QRMenuConnect was inspired by a simple need: restaurants should be able to offer fast, flexible access to their menus without expensive systems or long contracts. It came about when a potential client I tried to pitch for my text message loyalty and coupon service asked about a quick way menus customers can access that he changes several times a month instead of paying for a printed menu to be created, printed, and delivered—only to find he needed to change the menu again. I saw a need and searched for a simple solution. I found some already developed apps I could host on my servers, but they all had one thing in common: they didn’t make the solution simple and didn’t have a path to add more features to help the business owner. I wanted something that allowed the restaurant owner to create their menu and click a button to print a flyer that displays the menu’s QR code—table tent, counter stand, or window.
I designed it to print flyers in two practical sizes—4×6 postcards for table and countertop displays, and 8.5×11 letter size for windows and larger displays—each with the restaurant-specific QR code leading directly to a specific digital menu.
No more printing costs for menu changes. Just edit your QRMenuConnect menu and use the same QR code for customers to view your updated offerings. Understanding the importance of branding, I built in full customization: owners can edit colors, add their logo, dietary notices for each food item, activate/deactivate a menu, photo placement for each food item, and link their social media accounts to their digital menus.
Deployment was its own rollercoaster. Misunderstandings between me and the AI, tangled Firebase configurations, and versioning conflicts taught me more than I ever expected. But after countless iterations and mini-victories, QRMenuConnect.com finally went live on July 22, 2025, at 11:07 AM (version 1b)—a lightweight, affordable solution to help restaurants modernize how they connect with customers.
Beta Invite
Now, I’m looking for five passionate restaurant owners to join as beta testers. As part of this early group, they’ll receive free access to the Pro Plan, which includes powerful features beyond the Basic Free Plan:
- List up to 50 menu items per menu (instead of 20)
- Add up to 20 menu categories per menu (instead of 5)
- 4 active digital menus (instead of just one. Add images for each menu item
- Add social media profile links so customers can click icons on your digital menus
- Entitled to pro Plan features for free till beta offer ends (minimum 12 months from approval)
Their real-world feedback will help shape the future of the app. In return, I’ll spotlight their stories alongside mine—showing not just how the app evolved, but how every roadblock a bootstrap founder faces can become a lesson for anyone trying to build something from nothing.
Future versions are already in development and will offer much more—giving restaurant owners time to focus on what matters: running a great restaurant, rather than dealing with technology and marketing.
Example of one feature I’m working on: the ability to create a smaller-size QR code printout that can be ordered as stickers to place on takeout boxes.
And this is just the beginning of the bootstrap story of QRMenuConnect.com.
Final Words
To fellow developers, entrepreneurs, and those who are going through tough times: don’t give up.
I know what it’s like to lose everything familiar in your life. My career ended suddenly—not by choice or any of my actions—so I was forced to rebuild from scratch with limited means and a lot of uncertainty in my life and the outcome of the project. There were plenty of dark moments working on an idea. But coding this app, creating something real—something that can actually help small businesses—gave me purpose again. And yes, it’s not just a project, it’s a business. It’s a way to take control of your own destiny and financial future. Determination will set you apart from the ideas that were never tested to see if they’d work or not. The road for me has not ended in this venture; now is the time for traction building…
If you’re in the middle of your own storm, keep pushing. Even if it feels like everything’s stacked against you, progress—however small—is still progress. You never know when that late-night breakthrough or that one line of code will change your path. You don’t need a perfect plan, just persistence. It takes time, determination, and strength to build something great, rather than it being quicker to break something like someone’s career—or easier to give up on something or someone.
You are not alone, and your work matters—first to you, and then to others. It’s time to pull yourself up from your bootstraps and break through the roadblocks.
Go for throttle-up!
Aldo Grillo
