Is using a managed backend service a waste or a sensible choice when you already have backend skills?

Backend
1

I have been developing backends in the traditional way for several years. My work involves databases like PostgreSQL, setting up authentication, managing messaging queues, and ensuring system observability. This is part of my role at a tech startup focused on climate solutions, and I genuinely enjoy the challenge of integrating those components.

Recently, I’ve started working on some personal projects, which has led me to question whether it’s wise to delegate what I can do myself.

I’ve been testing out a platform called Gadget. To give you some background, Gadget provides a real PostgreSQL database along with built-in authentication and auto-generated REST and GraphQL APIs. While you can still implement custom backend logic using TypeScript or JavaScript, it simplifies many tasks that would typically take a long time to set up.

On one hand, I feel like I’m just paying for capabilities that I already possess. On the other hand, the reduction in time required is substantial. I can create prototypes in days rather than weeks. For instance:

  • I developed a journaling tool integrated with authentication and OpenAI support over a single weekend, which normally would have been a complex task involving user models, token management, rate limits, and background jobs.
  • I built an internal dashboard at my job that needed task queues and background processing, and Gadget managed retries and job tracking seamlessly.

So, I’d love to hear what this community thinks. If you have the skills to create everything yourself, but a service simplifies a large portion of the work, is using it just throwing money away? Or is it a practical move to spend your time on the unique aspects of your product?

I’m eager to hear insights from other backend developers, especially those who’ve navigated similar choices.

2

totally agree! if time is tight on personal proj, then use the tools you have. i mean, getting things done quickly can free you up for more creative stuff. but for prod, sticking to your own setup seems safer. just my 2 cents!

3

This really comes down to opportunity cost, not whether you can build it yourself. I’ve been building production systems from scratch for over a decade, and here’s what I’ve learned: the real value is what you build on top of the infrastructure, not the plumbing itself. My wake-up call was calculating how much time I actually spent on boilerplate setup versus features across projects. The ratio was embarrassing. Auth flows, database migrations, API scaffolding - I’d burn weeks on solved problems that added zero business value. Using something like Gadget doesn’t make you less of a backend developer. It makes you smarter about where you spend your time. Your system architecture knowledge actually helps you evaluate these platforms better. You’ll spot the limitations, understand what’s abstracted away, and know when you need to drop down to custom code. Your backend skills are still crucial for the complex, domain-specific stuff that actually differentiates your product. Let the platform handle the boring infrastructure while you tackle problems only you can solve.

4

hmm, curious about something though - have you hit any walls with gadget when doing complex backend stuff? what happens when you outgrow it or need custom database tweaks? feels like the real question isn’t whether you can build it, but whether you should be spending time on it right now.