I have a lot of Deep Learning/Machine Learning related pet projects; for some of them, I think that the public would use them (and maybe even pay for them). Of course, transforming a pet project into a SaaS is not one day of work (if I get this wrong please correct me).<p>My goal is to create a SaaS for one of my projects, and my main question is, what is a minimal architecture that I could follow to build something like that? And what tools should I use? Can you suggest books, blog posts on the topic?<p>If you created a SaaS, what are some of the things you would have known before you started?
I followed a similar path, turning an open source project I had started into a SaaS several years ago (it has since been acquired).<p>Two biggest things I learned: use whatever technology you’re most comfortable with and keep everything dead simple.<p>Your customers don’t care what language or framework you’re using. They will care when you can’t release features quickly enough because you’re stuck figuring out your own infrastructure.<p>I’ve written on HN previously[1] about some other things I’ve learned as well.<p>[1] <a href="https://news.ycombinator.com/item?id=34287685#34287996" rel="nofollow">https://news.ycombinator.com/item?id=34287685#34287996</a>
I'm working on a project that makes it easy to create Python front-ends: <a href="https://www.pycob.com/" rel="nofollow">https://www.pycob.com/</a><p>Do you have any examples of the pet projects posted on GitHub? I could try to take a stab at giving it a front-end, which would probably be the first step in making your pet projects SaaS
I used to use express.js with sqlite3. Can't get much more minimal than that.<p>Recently I tried remix.js and I think it looks quite promising. Most of the workflows I would use with express still apply, but the developer experience is so well thought out.