”Site reliability engineering”, or SRE for short, originates from Google around 2003. According to SRE inventor Ben Treynor, SRE is ”what happens when a software engineer is tasked with what used to be called operations.”
An SRE engineer is expected to do up to 50% ”ops”-related work (work on issues, be on call…) while the other 50% is for developing new features, automating things etc, so you need to be good at both system administration, coding and automation.
Read more: Site Reliability Engineering on Wikipedia
Read a book or two about SRE for free:
On the other hand, not everyone agrees that SRE is the way forward: Read about why error budgets might not always be a good idea when building resilient systems and teams: