Gen AI-Powered Q&A Feature

QUESTION	ANSWER
What content is included in the Q&A Feature?	We used Harvard Business Review (HBR) articles, podcasts, and videos in creating this feature. Not only is this content provided in recommended links but infused in the synthesized responses. We will constantly be pulling in the latest content from HBR as it is published.
Did you build the Large Language Model (LLM) that you are using?	No, we are using well-known models as our foundation and augmenting with our content using the Retrieval-Augmented Generation (RAG) Framework. The tool is designed with a modular architecture to allow easy updates to the language model or system prompts.
What is the Retrieval-Augmented Generation (RAG) Framework and how does it work?	RAG helps address LLM limitations by grounding responses in trusted, up-to-date content. When a learner asks a question, we search for relevant HBR content and send it, along with the question and system prompt, to the LLM. The LLM returns a synthesized response rooted in those resources, ensuring accuracy, currency, and transparency.
How is this feature different from what comes from other gen AI tools?	It offers reliable, unbiased, and HBR-styled answers. Safeguards prevent inappropriate responses. Learners also receive resource recommendations and progress tracking via their Leader Profile.
How does HBI store and use employee input and results generated through these AI capabilities?	User queries are stored and shown in chat history. Chats are not reported to the organization. Analytics use anonymized data only.
Can the gen AI functionality be deactivated based on company preferences?	Yes, it can be turned off at the organization level.
Is there a limit to the number of questions a user can ask the Gen AI-powered Q&A feature on HBR Spark?	No, there are currently no limits and no plans to introduce any.
Is it possible to implement guardrails to prevent learners from sharing proprietary or personal information?	No, users are advised not to input proprietary or personal information, but we cannot control what they choose to share.
Are there plans to expand the Q&A AI feature's knowledge beyond HBR Spark content?	No, it will remain focused on leadership and management within the HBR Spark platform.
Will reporting be available to identify common themes in AI inquiries or provide insights into trending content recommended by the Q&A feature within my organization?	No reporting plans are in place for this feature at this time.
Will the Terms of Use and Privacy Policy change once the feature is out of beta?	Terms of Use may be evaluated for changes post-beta. The Privacy Policy will remain unchanged.
Will you continue to enhance the Gen AI-powered Q&A Feature?	Yes, enhancements include streaming functionality, source improvements, follow-up question support, advanced classification, dynamic UI, and multilingual support.