* What you see *versus* what you get. The marketing, hype, tribe culture is strong. Everything looks great, but when you peel the onion there is surprisingly little substance/value. Almost $1 billion later and very little real-world operational value. The hardest part is knowing how little gets done and how much is spent. Case in point, every single week everyone is invited (and hundreds attend) a "wins" meeting to talk about all the amazing work done that week. This is probably a $40+k meeting. Maybe for some this helps the culture etc., but for me it's a distraction. Maybe bimonthly ftw ;)
* Meetings. For an agile organization there are a lot of them. Multiple daily standups (which often run long), pre-IPM, IPM, retro, health checks, team fun time, sync & integration meetings, trainings, and professional development. On top of all that you're pair programming 100% of the time. It has been an rough 2+ years.
* Transparency and communication with leadership is lacking. Poor decisions with no/minimal communication/conversations as far as which teams to standup, disband, staff-up/down, strategic technical priorities, etc. There are multiple teams that could go away with minimal consequences. KR often reinvents wheels in-house.
* Tooling is a mixed bag - some are awesome (e.g. Macs, GitLab), some are terrible (e.g. Pivotal Tracker isn't even in the top 10 bug trackers). Also, welcome to MFA/authentication joys - I have 5 different apps/mechanisms to login to various KR services at this point (e.g. Okta & Duo & KRID (in-house) & Google Authenticator (w/5+ different codes) & username/password) - no SSO here! Even hitting nexus requires MFA (more slowdown)
* After all the meetings it's also part of the culture to spend anywhere from 2-5x as much time writing, maintaining, and running tests than you get to spend writing new code. I spend 10-15% of my week pairing with someone writing new functionality. That's expensive and painfully slow progress when you have users that were engaged.