Compare Workflows
Compex organizes protected-data work into Train, Analyze, and Serve. The right workflow depends on whether you need to create a model, compute an answer, or run an approved service against protected data.
Train
Choose Train when the goal is to produce or improve a model using protected data under governance.
Analyze
Choose Analyze when the goal is to answer a bounded question, compute a score, or produce controlled analytical output.
Serve
Choose Serve when the goal is to expose a model, API, or application to approved users under policy.
What this means
Business teams should start by naming the output they need. If the output is a model artifact, start with Train. If it is a report or result, start with Analyze. If it is an ongoing service, start with Serve.
Current status
All three workflows are represented in the sandbox and can be scoped for pilots. The depth of implementation depends on the client environment and the selected scenario.
Limitations
Some projects span multiple workflows. For example, a team may Train a model, Analyze validation results, and later Serve the model to approved users. Treat the first pilot as a focused entry point, not the entire program.
See sandbox examples
Review synthetic examples that show how the three workflow types differ.