MLflow pyfunc serving is too slow for your SLO. What do you do while keeping the registry?
PICTURE THIS: HOW TO EXPLAIN IT
Simple meaning
Keep MLflow as the catalog and export the same version to Triton, ONNX Runtime, or a compiled custom server.
WHY — MLflow instead of guessing?
Why interviewers care about MLflow:
who only read docs from people who shipped.
and tied to MLOps work.
Name the idea, why it exists, then one short example.
End with when you use it and one common pitfall.
STEPS — What happens step by step?
Before you speak the answer, walk the interviewer through these steps:
- 1Keep MLflow as the
catalog and export the same version to Triton, ONNX Runtime, or a compiled custom server.
- 2CI would load both
and diff predictions on a golden set.
- 3The registry still records
the serving image digest so you do not lose lineage.
- 4Give an example
One tiny concrete case you can say aloud.
- 5Common mistake
What juniors usually get wrong.
- 6Close
When you pick this over the alternative.
EXAMPLE — See it in action
Here's a short line you can speak, broken into clear beats:
Note: Adapt this scaffold to your own project — keep it under 60–90 seconds.
Key takeaway
Keep MLflow as the catalog and export the same version to Triton, ONNX Runtime, or a compiled custom server. CI would load both and diff predictions on a golden set.