3
3 Comments

Please Validate my Idea: Launch wizard for AWS

Hello, I'm Victor. My day job is as a Sr Engineer at one of the FAANGs. I have a side hustle as a fractional CTO.

In my fractional CTO practice, I have noticed that many dev agencies still launch infrastructure with click-ops. It must take at least a few hours to get everything set up: launch ec2 instance, install php, point the dns name by hand, etc. And the end result is actually quite bad - stateful infrastructure that can't be easily replaced, where the DNS name is pointing directly to an ec2 instance. Often, they even install mysql on the same host.

All of these things are huge points of technical debt AND it takes longer to do it the wrong way than to do it the right way. My product would save dev agencies time and enable them to deliver a better solution.

My product vision is a website where you can launch new infrastructure into your (or your client's) AWS account in under 3 minutes. You'd go through a wizard where you put in the domain name, then configure your domain registrar with nameservers, then pick a couple more parameters, and at the end you have:

  • Scalability out of the box: an ec2 autoscaling group behind the dns name you want, with either Rails, laravel, or nodejs/express installed and running. The database would be on AWS RDS with multi-AZ failover for prod (but not for staging).
  • Source control: all the stuff i'll launch for you will live in a github repo. It will have a sample hello world endpoint and the infrastructure defined as code, using CDK. This repo lives in your own account or in your customer's account.
  • CI/CD out of the box: the repo will come with preconfigured github actions so when you push to the staging branch it pushes to staging environment, and when you push to the main branch it pushes to prod.
  • Several best practices: log pulling out of the box, log statements to cloudwatch out of the box, per-API metrics out of the box, and https out of the box, good load balancer configurations out of the box, database encryption at rest out of the box, etc.

You can come back to the website at any time and have a look at your apps. It would be like a catalog of apps you've launched and are maintaining for customers.

The initial (MVP) version of the platform would only allow you to see the list of apps you have launched and to launch new ones. There would be some deep links into cloudwatch for logs for each app but nothing else in terms of maintenance/observability.

Since the resources are all connected to your aws/github accounts, you can cancel at any time and keep running everything yourself.

My question is to the engineering side hustlers and consultants out there, do you think you'd pay $100/mo for this?

Even a simple yes/no would help. Thanks a lot for reading this far.

posted to Icon for group Ideas and Validation
Ideas and Validation
on June 18, 2023
  1. 2

    I have been in SaaS ecosystem building SaaS products, interviewing SaaS founders, running a newsletter for SaaS builders at Micro SaaS HQ, one of the biggest ecosystems for SaaS builders.

    The idea on paper might look good but finding a customers with such a need is almost impossible because of the complex combination of AWS, requirement of RDS.

    I understand most people go with a single EC2 instance with everything installed on one instance with no elastic IPs and no DR setup. But that is because of a reason where the business is more concentrating on marketing, sustaining in market rather spending money/efforts on complex tech. Agree DR is important but many not be the first priority. Agree that Elastic IPs are better but most people with a handful of customers wouldn't care unless there are hundreds of customers.

    Instead I would recommend you to build a SaaS starter kit and charge $500/yr for licensing/software updates. I have written about this niche earlier with a lot of examples.

    I would highly recommend you to recheck your idea and look at other niches too. I wrote about 100+ profitable niches in the last 2 years at Micro SaaS Ideas and that could be good starting point.

    Also, please look at FounderBeats where we interviewed founders building profitable products.

    1. 2

      Thanks! Where my idea came from is from the fact that it actually takes more time and work to go with the super crappy setup you described than to use a thing like what I had in mind.

      But after talking to a bunch of people I'm convinced you are right, people just don't care about this problem until way later. This business would be one of educating, and reaching the people who are open to being educated in this regard is impossible. Businesses where you have to educate are hard/bad too. Thanks for your input and the links!

      1. 1

        You saved yourself hundreds of hours. You are right - "This business would be one of educating, and reaching the people who are open to being educated in this regard is impossible"

Trending on Indie Hackers
I built a text-to-video AI in 30 days. User Avatar 50 comments I built something that helps founders turn user clicks into real change 🌱✨ User Avatar 49 comments You don't need to write the same thing again User Avatar 29 comments From a personal problem to a $1K MRR SaaS tool User Avatar 26 comments I built an Image-to-3D SaaS using Tencent's Hunyuan 3D AI User Avatar 25 comments What 300 Builders Taught Us at BTS About the Future of App Building User Avatar 24 comments