A collections team that receives call quality feedback once a month receives it too late to improve before the next examination. A collections team that receives call quality data weekly — individual agent scores, team averages, specific violation trends, and a ranked list of highest-priority coaching actions — receives it at the cadence at which behaviour actually changes. The Collections Training Agent AI produces a weekly QA report for the team supervisor every Monday morning: 5 minutes to read, actionable today, grounded in the actual calls from the previous week. It is the difference between a supervisor who manages impressions and intuitions about team performance and one who manages data.
A collections team that receives call quality feedback once a month receives it too late to improve before the next examination. A collections team that receives call quality data weekly — individual agent scores, team averages, specific violation trends, and a ranked list of highest-priority coaching actions — receives it at the cadence at which behaviour actually changes. The Collections Training Agent AI produces a weekly QA report for the team supervisor every Monday morning: 5 minutes to read, actionable today, grounded in the actual calls from the previous week. It is the difference between a supervisor who manages impressions and intuitions about team performance and one who manages data.
What a weekly QA report contains — and how it drives specific coaching actions
The weekly QA report is not a summary of aggregate metrics — it is a ranked action list. The supervisor reads it and knows, by the end of the first page, exactly which agents to coach this week, on which specific issue, and what outcome to expect. The report ranks agents by improvement trajectory (the agents improving fastest don't need weekly coaching; the agents declining fastest do), identifies the team's most common FPC risk phrase of the week, and flags any calls that produced a complaint or dispute in the week — linking the specific call, the specific phrase, and the specific outcome so the supervisor can address it concretely.
"A supervisor who reads a weekly QA report on Monday knows more about their team's compliance risk by 9 AM than they would know from attending 20 team meetings."
The weekly QA dashboard: week of November 11–14, 2025
Weekly Call QA Report — Collections Team · Nov 11–14, 2025 · Supervisor: Meenakshi R.
12 agents · 847 calls reviewed · 92.4% FPC compliant · 3 violation events · Week-on-week: +2.1pp compliance improvement
847Calls reviewed this week
92.4%FPC compliant (vs 90.3% last week)
3FPC violation events
4Agents requiring coaching this week
!
CA-152 · Anita Rajan · Priority coaching this week
FPC score: 68% (↓ from 79% last week)
Tone: 74%
Resolution rate: 38%
Anita's FPC compliance score fell 11pp this week — driven by the R02 violation (family contact offer) on the Nov 14 call with Vikram P., which also contaminated her week score. Additionally, 3 calls this week contained the phrase "I've noted that you refused to commit" — a high-escalation risk phrase that is technically compliant but correlates with a 41% higher borrower escalation rate. Her resolution rate of 38% is 14pp below the team average — linked to her pattern of dismissing partial-payment offers.
Coaching this week: 3 sessions (Mon, Wed, Fri) · Monday: R02 violation debrief and drill · Wednesday: partial payment acceptance script practice (role play) · Friday: escalation-risk phrase substitution ("I've noted you refused" → "let me note we're still exploring options"). Target: FPC score back to 79%+ by next week.
↑
CA-148 · Rohit Sharma · Improving · Continue R03 focus
FPC score: 74% (↑ from 66% last week, 58% two weeks ago)
Tone: 71% (↑ from 62%)
Resolution rate: 44%
Rohit is on a strong improvement trajectory — 16pp FPC improvement over 3 weeks, driven by daily R03 drills on prohibited language. This week, 0 R03 violations in 71 calls — a significant improvement from 3 violations in the equivalent period 3 weeks ago. His tone score improved significantly after the partial-payment script coaching session last week. Resolution rate at 44% is near the team average and improving.
Coaching this week: Continue daily R03 drills (not ready to reduce yet). Add R07 drill (penal rate disclosure) — Rohit correctly stated the rate on only 2 of 5 calls where the borrower asked. One brief coaching session Wednesday on penal rate calculation and disclosure. No full coaching session required this week — maintaining the improvement trajectory.
✓
CA-141 · Priya Iyer · Top performer · Peer coaching candidate
FPC score: 96% (↑ from 94%)
Tone: 88% (↑ from 85%)
Resolution rate: 62% (team high)
Priya is the team's highest performer across all three dimensions this week. Her 96% FPC compliance score represents 2 minor technical issues across 78 calls — both were borderline R01 edge cases (call ended at 7:01 PM, minor discrepancy). Her resolution rate of 62% is 10pp above the team average — driven by consistent partial-payment acceptance and commitment anchoring. The 3 calls Priya made to borrowers who had previously refused all other agents produced payment commitments in 2 of 3 cases.
This week: No remedial coaching needed. Meenakshi to listen to 2 of Priya's successful recovery calls — identify the specific phrases and techniques for use in team coaching library. Consider asking Priya to run a 20-minute script workshop for the team on partial-payment anchoring. Maintain R01 edge-case awareness (call-end timing).
→
CA-155 · Suresh Nair · Stable · Watch resolution rate
FPC score: 84% (→ flat)
Tone: 76% (→ flat)
Resolution rate: 31% (↓ from 39% last week)
Suresh's FPC compliance is stable and acceptable. The concern this week is the resolution rate decline — from 39% to 31% — which is 21pp below team average and Priya's 62%. Call analysis shows Suresh closing calls prematurely when borrowers express difficulty, rather than exploring partial commitments or follow-up call scheduling. He is FPC-compliant but leaving value on the table in every call where he encounters resistance.
This week: One coaching session focused on handling borrower resistance — specifically the "I can't pay" response and the partial payment exploration technique. Script: "I understand — if the full amount is difficult right now, what could you manage? Even a smaller amount would help." Suresh's FPC compliance does not need attention; his resolution technique does.
How the feedback loop compounds week-on-week: the 8-week improvement record
| Week | Team FPC compliance | FPC violations | Avg resolution rate | Key coaching action that week |
|---|
| Week 1 (Sep 16) | 81.2% | 11 | 39% | Baseline established · R01 and R03 most frequent violations · Daily drills started |
| Week 2 (Sep 23) | 83.8% | 8 | 41% | R01 drill frequency increased · 2 agents coached on call-timing awareness |
| Week 3 (Sep 30) | 85.4% | 7 | 42% | Escalation-risk phrase library built from week 2 calls · Shared with full team |
| Week 4 (Oct 7) | 87.1% | 6 | 44% | Partial payment acceptance script introduced · Priya Iyer used as peer model |
| Week 5 (Oct 14) | 88.6% | 5 | 46% | R02 drills added after 2 family-contact incidents · Agent CA-161 coaching completed |
| Week 6 (Oct 21) | 89.9% | 4 | 49% | Borrower resistance handling added to weekly drills · Commitment anchoring technique |
| Week 7 (Oct 28) | 90.3% | 4 | 50% | Priya's partial payment calls used in team coaching library · 4 agents adopted technique |
| Week 8 (Nov 11) | 92.4% | 3 | 52% | Ongoing · Anita R02 violation flagged · Target: 94% compliance, 54% resolution by Nov 25 |
+11.2ppFPC compliance improvement over 8 weeks — 81.2% → 92.4% · From weekly targeted coaching on specific violations · Not a one-time training event
−8FPC violations per week — 11 violations in Week 1 → 3 in Week 8 · 73% reduction · Each week's coaching targets the previous week's violations
+13ppResolution rate improvement — 39% → 52% over 8 weeks · Partial payment technique adopted by 9 of 12 agents · Priya Iyer as peer coach
MondayWeekly report delivered every Monday morning — supervisor reads in 5 minutes · Knows exactly who to coach and on what · No data gathering required
11 FPC violations in Week 1 became 3 in Week 8 — not because the team changed, but because the feedback loop identified which specific behaviours to target, and targeted them every week until they changed
The 8-week improvement record is not a story of better hiring or a different management style. It is the mechanical result of a feedback loop that operates every week: violations are identified, the specific phrases and situations that caused them are catalogued, drills targeting those specific scenarios are assigned to the agents who made them, and the following week's data shows whether the coaching worked. When Rohit's R03 drills moved his score from 58% to 74% over 3 weeks, the loop confirmed: more of the same. When Anita's FPC score fell 11pp in one week, the loop identified the cause, scheduled 3 coaching sessions, and set a recovery target. The supervisor reads this on Monday morning and knows what to do today — not because they are a better manager than last month, but because the system gives them better information than last month, at the cadence at which behaviour actually changes. The Collections Training Agent AI's weekly QA feedback loop is the mechanism that converts call quality measurement from a compliance document into a team performance tool — and the improvement compounds, because each week's coaching closes the gap the previous week's data identified.
