Hi, I am new. Does Kedro reduce inference latency? Thanks.

No. I'm not sure if you got any indication that it would?

If your pipeline structure is designed for multithreading or multiprocessing, then I suppose it could.

Depends on what you mean by inference latency. Does it reduce execution time of

model.predict()

for a single observation? No. Though if you for example have your observations split into multiple files, kedro can help parallelize inference on each of those. But single calls to

model.predict()

are not affected by Kedro.