Exactly! I have the same experience. It is a source of eternal frustration in the team. I had situations that everyone was crushing it, the team work efficiently and at the end the team had to go through the bashing on retro.
There is an argument that you can always add features to the sprint, so you start small and then extend sprint with additional task. But it does not work, because tasks seem to take all the time that was planned for them.
However, the rituals of SCRUM like retros and groomings taking place every 2-3 weeks is a good thing. It's like hygiene.
That's very good and any ISO 270001 certified small-medium company would need that. But it would be better to have stand alone web interface and not only relay on Slack.
If you are far behind start with Khan Academy. Then there are MOOCs on Coursera and eDX as well. Less interactive but 100% Real deal (and it doesnt mean harder) are course materials often with videos from MIT Open CourseWare. If you are into ML do Gilbert Strang Linear Algebra. It will develop you so much that it can’t be understated. Buy his book and do exams and assignments. If you want to take pleasure in math there are brilliant channels on yt, for example Numberphile. GL man.
The site is served via AWS CloudFront. It is SPA app hosted on S3. The API is on Heroku. No mem-cache anywhere. To search we use ElasticSearch hosted on AWS. We bump up the specs to speed it up but that was not the core problem.
I suspect Heroku Postgres DB. We didn't used some Hobby version and I think it is the real bottle neck + some unnecessary middleware for the API that also requires DB.
It's nice to meet competitors. We have run the name against trademarks databases. The product was not ready when we registered our domain, but that's valuable feedback, to look out for similar domains.
We had API cached by ClodFront. In the process of troubleshooting we turned it off and we attached frontend directly to API. There should be CORS headers though. The problem was more scalability problem. It's too soon to say for sure, but getting rid of some middleware and scaling Heroku Dynos helped.
We run it on Heroku on production quality dynos. Heroku came out to be a bottle neck. Moving things to lambda is an option we will seriously consider. Checking for, Twitter as well.