Once your experiment is live, the assistant shifts from setup to interpretation. Open it from the assistant icon in the top right of the experiment — it can see the experiment's configuration and status, so your questions are answered in context.
It Explains the Recommendations
When the optimizer suggests the next conditions to try, the assistant tells you what they are exploring and why they are worth running — in specific terms, not generic commentary.
Ask things like:
"Why is it suggesting such a high temperature this round?"
"What is this batch trying to learn?"
The assistant fetches the recommended conditions and explains the reasoning behind them, so you can run each round with an understanding of what it contributes to the search.
It Refines the Guidance with You
As your understanding of the problem evolves, the assistant updates the experiment's expert context, name, or description — so the optimizer's steer always reflects what you know now.
For example, if early rounds reveal a relationship you hadn't captured — "it turns out baking time matters much more above 200 °C" — tell the assistant, and it will work the observation into the expert context with you.
What It Won't Do on a Running Experiment
A running experiment keeps its shape. These boundaries are deliberate:
It does not change the structure. Parameters, constraints, and the algorithm of a running experiment remain under your control in the configuration view. If you want to make such a change, the assistant won't do it for you, but it will guide you to where and how to make it yourself.
It does not enter your measurements. The assistant never submits measurement values on your behalf, and it never invents them. Your results are yours to enter.
It does not restart or retrain. It tells you when a model refresh would help; the action is your explicit click.
What It Can and Can't See
The assistant reasons from the experiment's data — its configuration, status, and results. It does not yet see your charts and plots, so if a question is about something visible only in a visualization, describe what you're seeing and the assistant will reason from the underlying numbers.
Related Articles
Meet Your SDLabs AI Assistant — overview and guardrails.
Designing an Experiment with the AI Assistant — the setup journey.
Starting from an Existing Experiment Description — paste a write-up to begin.