Improve AI Coach Context Awareness and Conversational Adaptability | Voters

Improve AI Coach Context Awareness and Conversational Adaptability

under review

Himanshu Bishwash

The AI coach should better recognize user-provided context (e.g., long-term training habits, goals, calibration concerns) and adjust its guidance accordingly. It should avoid repetitive static replies and demonstrate adaptive reasoning to improve trust and engagement.

3 days ago

Amanda

marked this post as

under review

Hi Himanshu Bishwash, thank you for the feedback. Have you tried telling Bevel Intelligence all of your goals or concerns? The daily messages will use this context accordingly. Do you have any screenshots of what you've said, and what you'd expect Intelligence to say? This would be very helpful. Thanks!

Himanshu Bishwash

Amanda It took me a while to reach a point where the Coach prompted me to provide this data before which, the Coach would just provide ‘Strong Recommendations’ and argue that it is accurate. Despite me explicitly suggesting that it recalibrate it’s baseline, the Coach would dismiss the requests, citing high confidence in Bevel’s backend formula.

I have long and multiple chats with the Coach. It costed me my sleep and ruined my next day, but that’s on me. I have attached an example. This chat went to length after this, but the responses were same and repetitive. Instead, the a Coach should have asked for further context and tried to considerate rather than being rigid and dismissive.