@RR5aF0V-1gxh as a software enigeneer know this: Juniper's product development cycle is one of the longest, if not the longest in the industry. The biggest reason why the current trend among Juniper dev team is to leave enmass for Arista. Well, you can too after you are there for a while and figure out that showing one release of CSO in your CV, which does not work anyway in 1 year cannot be the greatest career move you have ever made.
For a starter, Juniper pay package is good, as it was in my case; the first one odd year was also good and then I realised that people who stayed there were either career Juniper people, worked their for 7+ years or they leave - I am leaving, I am not a career Juniper drone. The product dev thought process is a mess. They announced QFX5300 to customers, scrapped it; they started talking about SD-WAN, CSO etc. in Nov 2015, CSO is now version 3.2 and still scratchy - and Cisco now has ENCS, Viptela and NSO (Tail-F) - all field tested except for ENCS but that's a box and that will be good.
And on top of all that there is the sword of Democles hanging on everybody - poor management, spending too much time , money and energy behind non-existent solutions (SDSN, IoT,...), stock price at a freefall after each QBR..the list goes on. They have ppt's wih SDSN all over BUT for SDSN solution tey have SRX and vSRX and that's it - rest are BS wrapped in "automation", "orchestration" , "ML" - all the key industry words but no product/solution relevance They will be firing people - now whether or not they wake up and fire the guilty or protect the guilty as they always have and fire the sacrificial technical is something to be seen.
If I were you, I'd hang around to join a Arista or a Cisco (yes, their software strategy is completely revamped and getting traction from customer base as reflected in stock price) or a Cumulus Networks (who are looking for people, I am just off my fourth interview and amazing solution) . Whatever you decide, best wishes staying out of the swamp.