A user asks watsonx Assistant 'What are the onboarding requirements?' and then follows up with 'How long does the process take?' The assistant fails to connect the follow-up to the earlier topic and returns an unrelated answer. What should the engineer configure to resolve this?

For coherent multi-turn conversations, the search integration must be configured to pass previous conversation turns to the LLM as part of the context window. Without this, each user message is treated as an isolated query and the model cannot resolve pronouns or implicit references back to earlier turns in the session. The correct answer, "Enable session history in the search integration so prior conversation turns are included in the LLM context window", directly satisfies the scenario because it aligns with watsonx Assistant's design principles and the specific capability being tested. The incorrect options ("Add more NLU training examples for each type of follow-up question", "Create a separate action for every anticipated follow-up question type", "Increase the dialog node confidence threshold to force more specific intent matches") may appear relevant but each misses a key requirement or introduces a step that is either unnecessary or belongs to a different workflow. Mastering the distinction between these approaches is essential for effective watsonx Assistant implementations and is a core focus of the Domain 4: Integrate with watsonx section of the certification exam.