1
00:00:00,400 --> 00:00:04,620
A relief delivery is only part of the job. The coordinator still

2
00:00:04,620 --> 00:00:08,488
needs the receipts, the counts, and a report for every donor.

3
00:00:09,600 --> 00:00:15,292
BasketBrief helps small relief teams turn scattered evidence into approved donor reports.

4
00:00:15,292 --> 00:00:17,664
Here is one complete example.

5
00:00:18,767 --> 00:00:24,095
Meet Amal, our fictional coordinator. Rana has the receipts. Sami tracks deliveries.

6
00:00:24,095 --> 00:00:27,647
Amal needs a consistent report for two donors.

7
00:00:28,767 --> 00:00:33,255
The agent asks contributors for missing evidence and prepares the report. Amal

8
00:00:33,255 --> 00:00:37,743
reviews it and decides what gets shared. Let us follow that process.

9
00:00:38,900 --> 00:00:43,639
The team reports twelve hundred and sixty dollars spent. Sixty dollars has

10
00:00:43,639 --> 00:00:47,588
no receipt. BasketBrief reads the sources and finds that gap.

11
00:00:48,900 --> 00:00:52,773
Here is Rana’s inbox. The agent has asked her for the transport

12
00:00:52,773 --> 00:00:56,646
receipt. Amal does not have to relay the question, and the same

13
00:00:56,646 --> 00:00:58,260
answer will support both reports.

14
00:01:00,900 --> 00:01:05,593
Rana uploads the photograph. Nova Pro reads it through AgentCore Runtime. The

15
00:01:05,593 --> 00:01:10,287
original and the transcription stay together so the coordinator can check what

16
00:01:10,287 --> 00:01:11,460
the machine read.

17
00:01:12,900 --> 00:01:17,460
The evidence is ready. Amal reviews the report and approves this version.

18
00:01:17,460 --> 00:01:22,020
The agent can prepare and follow up; approval stays with the coordinator.

19
00:01:23,900 --> 00:01:28,741
The first report is now in the donor’s inbox. It records ninety-two

20
00:01:28,741 --> 00:01:32,372
delivered and eight returned. That copy will stay intact.

21
00:01:33,900 --> 00:01:38,469
After these reports are sent, Sami updates the delivered count from ninety-two

22
00:01:38,469 --> 00:01:42,276
kits to eighty-eight. The team now needs an amended report.

23
00:01:43,433 --> 00:01:48,805
Sami reports eighty-eight delivered. Four kits no longer reconcile. BasketBrief stops the

24
00:01:48,805 --> 00:01:53,729
earlier approval from authorizing this changed report, and opens a follow-up.

25
00:01:55,433 --> 00:01:59,673
This is the question in Sami’s inbox. The system keeps households unknown.

26
00:01:59,673 --> 00:02:03,912
A difference in kit counts cannot tell us how many people received

27
00:02:03,912 --> 00:02:04,265
help.

28
00:02:06,433 --> 00:02:11,718
Sami checks storage and confirms twelve returned. The answer closes the gap.

29
00:02:11,718 --> 00:02:15,241
BasketBrief prepares the amendment from the updated evidence.

30
00:02:16,433 --> 00:02:21,531
Amal sees both changes before approving: delivered falls to eighty-eight, returned rises

31
00:02:21,531 --> 00:02:24,929
to twelve. The amendment needs its own approval.

32
00:02:26,433 --> 00:02:31,397
Northstar receives the approved amendment. The exact changes are visible, and the

33
00:02:31,397 --> 00:02:33,465
original report is still there.

34
00:02:35,433 --> 00:02:40,482
The Arabic donor receives the same facts and amendment. One evidence trail,

35
00:02:40,482 --> 00:02:42,585
carried through to both recipients.

36
00:02:43,700 --> 00:02:48,546
For your own team, each person signs in separately. New evidence creates

37
00:02:48,546 --> 00:02:52,988
a notification for the responsible reviewer. Donors see approved reports only.

38
00:02:54,700 --> 00:02:59,846
A requested receipt attaches to the original expense, avoiding a second charge.

39
00:02:59,846 --> 00:03:04,564
Multiple currencies stay separate until a reviewer supplies an exchange rate.

40
00:03:05,667 --> 00:03:12,219
Strands and Nova interpret messages and choose tools. AgentCore Runtime reads images;

41
00:03:12,219 --> 00:03:18,771
Memory provides advisory vendor history. Code enforces arithmetic, permissions and versioned approval.

42
00:03:20,000 --> 00:03:24,534
We tested the complete live loop, separate team accounts, and a private

43
00:03:24,534 --> 00:03:29,067
Gmail attachment import. Our automated tests cover the boundaries. Field impact and

44
00:03:29,067 --> 00:03:31,712
time saved have not yet been measured.

45
00:03:34,000 --> 00:03:38,696
From the first missing receipt to the last donor update, BasketBrief keeps

46
00:03:38,696 --> 00:03:43,000
the follow-up connected, with a person in control of every approval.