ABCDEFGHIJKLMNOPQRSTUVWXYZ
1
AR Calling Automation — Demo Tracker
2
Built for: Cloudify Books & Advisors (accounting firm demo)
3
Demo as-of date: 2026-05-17 — change in Config!B4 to re-anchor.
4
5
WHAT THIS TRACKER DOES
6
This workbook is the data backbone for an AI-powered AR collections system. It tracks overdue invoices, runs an automated multi-channel follow-up cadence (email → SMS → AI voice → human → pre-legal), captures Promises-to-Pay (P2P) from AI call analysis, and feeds a live dashboard. Export to Google Sheets to plug into n8n / Zapier and Vapi.
7
8
SHEETS OVERVIEW
9
• Dashboard — KPIs, aging buckets, today's action items.
• Invoices — master AR list. Aging, stage, and next-follow-up are formula-driven from As_Of_Date.
• Customers — clients with risk tier, preferred channel, best call window.
• Call_Log — every outreach attempt with AI analysis written back from Vapi → Claude.
• Promises_to_Pay — commitments captured; auto-updates honor rate.
• Follow_Up_Cadence — rule engine (stages, channels, frequency, escalation).
• Payment_Patterns — per-customer learned behavior used by the AI agent.
• Config — As_Of_Date, dropdown lists, named ranges.
10
11
HOW THE AUTOMATION FLOWS (n8n)
12
1. Trigger: daily 8:00 AM cron in n8n.
2. Read Invoices sheet → filter to rows where Next_Follow_Up_Date <= today AND Status not in (Paid, Disputed).
3. For each row → look up Customer + Pattern + Cadence rule.
4. Compose channel-appropriate outreach (email, SMS, or Vapi voice).
5. For voice: kick off Vapi call with stage-specific assistant prompt.
6. Webhook receives call transcript → Claude extracts P2P date/amount, sentiment, dispute, next action.
7. Append row to Call_Log; if P2P captured → append to Promises_to_Pay.
8. Update Invoices.Last_Contact_Date and Invoices.Next_Follow_Up_Date (or use formula).
9. Nightly: bank-feed reconciliation marks P2Ps Honored / Broken; updates patterns.
13
14
AS-OF DATE
15
All aging/stage/next-follow-up formulas reference the named cell As_Of_Date (Config!B4 = 2026-05-17). Change it once and every formula re-anchors. Useful for re-running the demo without editing 28 invoice dates.
16
17
DEMO NARRATIVE TO TELL BUYERS
18
1. Show the dashboard: $111K AR, 11 active P2Ps, 1 broken, 2 disputes, 2 pre-legal. 'Yesterday this was a partner-time problem; now AI handles tier 1-2 and the partner only sees tier 3+.'
2. Show Invoices sorted by stage: 'Each row already has its next channel, time, and tone decided.'
3. Open Call_Log row 5 (CALL-001): 'Vapi made this call in 2 min 22 sec. Claude wrote this summary, extracted the P2P, and flagged it for confirmation on the promise date.'
4. Show the BROKEN P2P (CALL-010 / INV-0057): 'When the customer didn't pay on commitment, the cadence auto-escalated.'
5. Show the dispute (CALL-002): 'AI agent recognized scope dispute, didn't push payment, escalated to partner.'
6. Show Payment_Patterns: 'The system learns each client. SMS works for C-008, voice for C-005, partner for C-003.'
19
20
COLOR LEGEND
21
Aging buckets: green=current, yellow=1-30, orange=31-60, light red=61-90, dark red=90+.
P2P status: green=honored, blue=pending, red=broken.
Sentiment: green=positive, yellow=neutral, red=negative.
Risk tier: green=low, yellow=medium, red=high.
Blue text = hardcoded input you can change. Yellow fill = key assumption.
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100