Does AI triage still save time after review?
Check a fictional AI memo against routing, human review and fixed support time.
The question
Harbor Claims is a fictional US service operation that routes 1,000 claims each week. This exercise concerns staff time for routing, not approval of any claim. Complex claims stay manual. The supplied review time meets the exercise’s quality requirement.
| Value | |
|---|---|
| Claims | 1000 |
| Simple share | 60% |
| Manual minutes per claim | 8 |
| AI handling minutes per simple claim | 2 |
| Review minutes per simple claim | 3 |
| Fixed support hours per week | 20 |
Complex claims stay manual. No cash savings or staffing reduction is assumed.
The 600 eligible simple claims each save six minutes. This releases 60 staff-hours per week.
“How much staff time does the targeted plan save after review and support? Recheck it when the simple share falls to 30%.”
Not saved by this site. Your browser may restore text when you navigate back.
A strong answer, step by step
Open each step in turn. This is one strong answer, not the only one.
Step 1: Which parts of the memo survive a check?
What would you do here? Open the move you’d make.
Reject the memo because it applies AI to complex claims.
Tempting, but it misses something. The memo explicitly uses only eligible simple claims.
Keep its routing and arithmetic, then add the omitted work.
The strong move. This identifies the actual error without changing a valid subtotal.
Accept 60 hours because its arithmetic is right.
Tempting, but it misses something. The calculation does not include review or fixed support.
See how a strong candidate says it
“The memo routes only the 600 eligible simple claims and calculates 60 hours correctly before review. It leaves out three required review minutes per eligible claim and 20 fixed support hours.”
Why it works: An accurate subtotal can mislead when work is missing from the scope.
Step 2: What is the net weekly time saving?
What would you do here? Open the move you’d make.
Subtract only 20 fixed support hours from the memo’s 60.
Tempting, but it misses something. That still omits 30 hours of required human review.
Subtract only review from the memo’s 60 hours.
Tempting, but it misses something. That gives gross savings but ignores fixed support.
Include review, convert to hours, and subtract support.
The strong move. It accounts for each type of staff work in the same unit.
See how a strong candidate says it
“Manual: 1,000 × 8 = 8,000 minutes. Targeted: 600 × (2 + 3) + 400 × 8 = 6,200 minutes. Gross saving 1,800 minutes = 30 hours. Subtract 20 support hours for 10 net staff-hours per week.”
Why it works: Separate handling minutes from weekly support hours before subtracting.
Step 3: What decision does the initial result support?
What would you do here? Open the move you’d make.
Pilot the simple-claim path and verify time and review quality.
The strong move. This ties the limited recommendation to the supplied assumptions.
Roll out to every claim because the initial net result is positive.
Tempting, but it misses something. The estimate applies to simple claims only; complex claims stay manual.
Remove human review to raise savings.
Tempting, but it misses something. Required review is a constraint, not a discretionary cost.
See how a strong candidate says it
“I would run a targeted pilot if the review time and quality requirement hold. Ten net hours per week gives a reason to test the workflow. I would verify the actual mix and measure the time before a permanent rollout.”
Why it works: A defer-and-measure choice is also defensible if the review or mix inputs are unverified.
Step 4: Simple share falls to 30%. What changes?
What would you do here? Open the move you’d make.
Keep the initial recommendation because 1,000 claims still arrive.
Tempting, but it misses something. Total volume is unchanged but the share that can save time is lower.
Recompute time saved on 300 eligible claims and retain support cost.
The strong move. The changed mix reduces gross savings while the fixed support still applies.
Reduce fixed support to 10 hours because eligible volume halves.
Tempting, but it misses something. The exercise states that support is fixed. Changing it requires new evidence.
See how a strong candidate says it
“Only 300 claims now save three minutes each after review. Gross saving is 900 minutes = 15 hours. After 20 hours of support the result is −5 staff-hours/week. Keep manual routing for the stated objective, or get fixed support below 15 hours before rollout.”
Why it works: At 40% simple claims the plan breaks even. A positive time saving needs a higher share or lower support.
The answer, in short
- Decision
- Keep manual routing at the changed mix.
- Because
- The targeted path now uses five more staff-hours per week after support.
- What would change it
- A simple share above 40% or support below 15 hours at the changed mix would make the time case positive.
Common traps
- Treating a correct subtotal as the total.
- Leaving out required human review.
- Assuming fixed support scales with eligible volume.
Check your answer
- Did you identify what the memo did correctly?
- Did you include both omitted terms?
- Did you use staff-hours per week for the final answer?
- Did you update the decision when the eligible mix changed?
Last reviewed Nov 11, 2026