My highest engineering priority right now is implementing logging/alerting/incident response on critical systems so we can grant engineers elevated permissions while ensuring that an independent security organization can audit/monitor/remove access if credentials appear to be leaked or misused. This is another way of improving velocity through organizational resiliency. Also working on a number of new service designs and refactoring old systems.

My production environment is filled with tech debt, my team is afraid to change the code and our productivity is grinding to a halt. Can I save this without starting from scratch?

Yes! If you're ready to decide that you want to rid yourself of these shackles (and i generally believe you do to maintain a high-caliber environment) then you're going to need to start by establishing an expectation of automation around changes and deployments and exercise them on a regular basis. Along with the mantra of "always leaving things better than you found them" and educating engineers that deployments are always safe, repeating these deployments will incrementally turn a fearful system into one that you can safely make large changes.