I am always wondering how can a data engineer know, and then convince product team or business that their architecture or design is correct? There is not exactly an exact science to data engineering. There are companies which are happy using out of the box tools and make it work, and there are engineers putting up with complex legacy codebases using microservices et al. Besides reading up and doing courses trying to get as close as possible to best practices, your architecture suggestions are subject to either lessons learnt from experience (takes 1-2 years for a big data project to take real shape) or googling and doing the best possible back of the envelope tradeoffs. While other non technical team members expect you to give them a silver bullet or quickly lose trust in you. How do you establish this confidence in a solution without seeming like a tinkerer?