The trouble is that these positions try to filter you out asking for the same concepts in interviews. Nobody cares for how good a programmer you are. Data engineer positions don't work well if you want to touch cutting-edge concepts like big-scale DNNs. Sure you'd learn about deploying things but if your day-to-day job is not about e.g. ONNX, pytorch, JAX and massive scale training/inference pipelines, then you'd still be left out.

Chicken and egg problem. Probably the optimal way is just aiming for top ML research teams right from the start as a new grad. No other way is better.