Here's some advice from someone who has built (and currently lives off) a steadily growing and profitable solo SaaS business:
Here's what I use and what I think about it:
* Hetzner for servers, and use bare-metal ones, the speed and memory per dollar advantage over things like AWS is so large it's not even funny.
* Cloudflare for hosting domains and running your DNS (great).
* Braintree for subscription billing. It's not good at all, but Stripe is significantly more expensive and doesn't really get me that much more (it still can't handle EU invoicing with SAF-T export and its idea of invoicing is very US-centric). If you look at Stripe pricing and you are not in the US, look carefully: they will not deposit USD into a non-US account, which means they will hit you with currency conversion fees and poor rates. Add up all the fees and rates and you end up with 5.4% (last I checked).
* No ads. I stopped burning money on them after implementing my own tracking and finding out that I get exactly 0 signups through ads (tried Google, Facebook, LinkedIn, Quora).
* Linear for bug tracking. Fantastic tool.
* ProfitWell for tracking your subscription billing metrics.
* I still pay for Sendgrid, but I'd recommend against using them. They will send your E-mails from the same servers that their spammer customers use, so you will get plenty of rejected mail. No way to get around that unless you pay them big $$$ for dedicated IPs etc — it's a form of ransom, really. I send transactional mail myself, and for newsletters I'm looking for something better.
* Clojure + ClojureScript for software. Use a single language for both client and server, use the same business logic code for both, minimize line count, minimize programmer effort. An obvious bet for a solo founder.
* Ansible for managing your sever clusters, terraform for quickly spinning up experimental environments, and don't use AWS or heaven forbid Azure for those, use Digital Ocean which makes things really simple and saves you lots of time. Vultr is good, too.
That's it for tools, I think. But there is one thing I found more important than tools: I believe you should disregard most "common knowledge". Do not follow the hype. Read HN comments very critically: most people here are not in your situation. You need to optimize for different things than most HN readers. You are responsible for everything, including the bottom line of your business. So think for yourself. Don't jump into something just because lots of people write about it (ahem, Kubernetes). Don't do something a certain way just because it's current fashion (ahem, microservices). Don't use services just because most people do (ahem, AWS and Stripe). In each case, consider each service/tool carefully in the context of your business, your metrics and your requirements.
In my case, I am primarily optimizing for my time. But not only — I am willing to do some things manually (invoiced billing with wire transfers) or use a lower-tech provider (Braintree rather than Stripe) when it makes financial sense.
You mentioned PR and marketing — I also thought I would need to hire people, run campaigns, etc. I listened to all the podcasts and conference talks about marketing tools and strategies. And then I realized that most of these people do not run my type of SaaS — in fact, most of them make marketing tools for marketers. It's a huge echo chamber. I found out that with limited content marketing (e.g. writing articles from time to time) and "organic" spread, I'm getting a consistent number of signups. Could I get more? Yes, very likely so. But at what expense? Would these customers stick around? And would I be able to onboard and support them? So here again, think for yourself.
Unless of course your business is the same as everybody else's and you are building another tool for drip email campaigns, conversion tracking, idea voting, etc :-)