SLOs and SLIs are the core concepts of any SRE team. They allow us to keep the velocity of change high while balancing a platforms reliability. They are similar to SLAs but fit better in an engineering context
SLOs are designed to provide a certain level of service, defined using SLIs.
SLIs are ways for us to communicate quantitative data about systems. To agree that an SLO is met.
Is a way of using an SLO to define the capacity of error. For example, if we have an SLO stating: "95% of emails will successfully pass through FileTrust" then every 100 emails we have an error budget of 5 emails. You can use an error budget to control the velocity of releases. For example, if the success rate is 99% and you have a 4% budget left so you could increase releases that could increase the risk of the budget shrinking. In inverse, if the success rate is 80% due to increased releases that are breaking the system then you slow down releases till you have the budget.
Examples can be found on the SRE workbook