[D] Solo machine learning engineer woes

I work at a start-up as the sole machine learning engineer, and I have been tasked to develop a state-of-the-art model on a domain specific problem.

Some information on the task:

• Computer vision problem - complex keypoint localisation / 3D mesh prediction
• No training data - I'm proposing that I build a synthetic data pipeline supplemented with real-world data.

I feel like there is too much for me to be doing at once: reading papers, writing plans and documentation, implementing …

learning machine machine learning machinelearning machine learning engineer