AI Helper Effectiveness Summary

Positive Outcomes (Thumbs Up): 42 (78%)

Negative Outcomes (Thumbs Down): 12 (22%)

Most Helpful Offset: 0 (during task)

Least Helpful Offset: +2 (after completion)

“Offset” reflects when the Hope Strategic Assistant was invoked relative to the caseworker’s workflow: –2/–1 = before action, 0 = during, +1/+2 = after. These values help analyze whether the assistant supports decision-making or simply confirms work already done.

Outcome Ratings

Offsets for Successful Responses

Offsets for Unhelpful Responses

Reliability Trend Over Time

Tracks how often the Hope Strategic Assistant produced guidance that caseworkers rated as accurate, aligned with RCWs/FFPSA, and procedurally useful.

Total Questions Answered: 126

Questions with AI Assist: 74 (59%)

Avg AI Assist Position: +1.2 (-2 = well before, +2 = after)

Average Score Change (with AI): +11 pts

What’s Next: Review questions where AI changed your answer (9 items flagged)

New! AI Usage Insights beta is now enabled for this pod.

AI Usage by Day

AI Assist Offset Distribution (-2 to +2)

Offset is measured relative to when the question was first shown: -2 / -1 = before attempting, 0 = during, +1 / +2 = after initial answer.