I still remember the excitement of my scrum master training more than a decade ago in New York City.
With my new classmates I practised the ball point game and in groups, we built paper prototypes at remarkable speed. After the course, I went back to the software consultancy I worked for and gave the engineers a mini class on what we had covered (a deal I made with the CEO in return for paying for my training!) Both in the training class and back in my consultancy, you could feel the optimism of an approach that removed the barriers between users and engineers and favoured agile speed & iteration above requirements documentation. It was tremendously liberating for everyone to have a say on what would be built and for product and engineering to discover the best solutions, which considered both the user experience and the technical difficulty to implement.
Now in the present day, I believe there's a significant risk of breaking this pattern where teams have both a Product Manager and a Designer. Without a new feature planning process defined, it's all too easy for the PM & Designer to come up with new features together and only involve the engineering team at estimation time. This misses out on the initial brainstorming & ideation process which is so valuable. Both because it can provide novel solutions, but also because it gives the engineers a sense of ownership.
If you only involve the engineering team once the feature has been designed it's much harder for them to propose alternative implementations that could radically reduce the effort required and/or provide a better solution. In fact, you've inadvertently created a waterfall process!
My advice then, is to make sure that you define how you will work together across product, design and engineering during the ideation process and avoid creating an impediment to your agile development