Packs
Groups of talks organized by topic.
The model works when you run the notebook.
The score is in a spreadsheet, the model is in a bucket, and nobody can identify the data or environment that produced either.
A table loads successfully and still gives the model the wrong evidence.
Two models use a feature with the same name, but different calculations.
The preprocessing finished, but training never started.
Your code tests pass and the new model scores better, but an important customer scenario has got worse.
A rollback restores yesterday's model weights behind today's incompatible serving code.
A model that predicts quickly on your laptop can still leave users waiting.
The endpoint is healthy, yet recommendations are empty or one customer group receives worse predictions.
Production errors keep arriving, but the next model trains on the same old examples.
Adding GPUs made training more expensive, but barely shortened the run.
The bill is growing, but buying cheaper machines is only one possible response.
Nobody can name all the models in production.
A downloaded model brings executable code into your environment.
You built the platform, but data scientists still run their own scripts and ask colleagues how to deploy.
The prototype works, but nobody budgeted for the years of upgrades behind it.
The model is ready, but the application team has no time to integrate it.
The answer is somewhere in your documents, but retrieval misses it.
A prompt change looks better on your favorite examples but annoys real users.
An agent calls the right API but changes the wrong record.
The model runs on your GPU, but concurrent requests exhaust its memory.
The fine-tuned model sounds more like your examples, but the business decision is no better.