We're also experiencing very low quality responses. What's most frustrating to me is that they are completely non-deterministic. Using the SAME prompt when opening two different new blank chats (no previous conversation) results in different answers! So I cannot even configure the prompt to fix the response....
Hi @Arin Mak —Rovo’s accuracy depends not just on the prompt but also on the clarity and structure of the content it draws from. If your financial products have overlapping names or unclear metadata, the AI may get confused. Precise prompts and clean data help. Here’s a great prompting guide from from Atlassian - Ian Buchanan to improve results. Also remember to create a feedback loop with Rovo by rating response with the thumbs up and down icon.
Consistency and reliability are active areas of improvement for Rovo.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.