A product manager wants to send one prompt to several foundation models in a browser, read the responses side by side, and then choose a model to build on. Which Vertex AI surface is intended for this?
Show answer and explanation
Answer: C
Vertex AI Studio is the interactive console for prompt design and model evaluation. It lets users try prompts against foundation models and compare outputs, which helps with model selection before any application code is written. Cloud Shell is a command-line environment, ML.PREDICT runs inference on tabular data in SQL, and Composer schedules workflows rather than comparing model responses.
Why the other options are wrong
- A. Cloud Shell provides a terminal with command-line tools. It is useful for scripting but offers no side-by-side view of model responses.
- B. ML.PREDICT scores rows with trained BigQuery ML models. It is not built for trying prompts against foundation models and comparing text outputs.
- D. Composer orchestrates workflows as DAGs. It can call APIs on a schedule, but it is not an interactive environment for comparing model answers.