TF DNN Model
Volaris Processor Selection
Performance Dashboard
Real-time approval rate monitoring across WORLDPAY, ELAVON, MIT_BULK, and AMEX processors. Data sourced from Snowflake PAYMENT_ML.
Total Attempts (30d)
4,626,864
Weighted Approval
82.7%
Processors
4
Analysis Window
30 Days
2026-03-01 to 2026-03-31
◆Processor Overview
Approval rates and volume across all four processors. Apr 15 - Aug 14, 2026 (122 days)
AMEX
91.38%
336,326 attempts
-0.59 pp vs prev week
MIT_BULK
82.61%
2,347,975 attempts
+0.11 pp vs prev week
ELAVON
89.00%
127,635 attempts
-0.34 pp vs prev week
WORLDPAY
80.78%
1,814,928 attempts
+0.28 pp vs prev week
Volume Distribution | Mar 8 - Apr 8
Approval Rate Comparison | Mar 8 - Apr 8
◎Recommendations
Actionable items to improve model performance.
| Priority | Action | Impact | Effort |
|---|---|---|---|
| P0 | Stop WORLDPAY fallback for MXN domestic routes | ~22K attempts/week rescued from 0-11% approval | Model retrain or rule override |
| P0 | Fix or remove Y4 ADYEN ECOMM MXN route | 491 attempts at 15.89% approval | Routing rule change |
| P1 | Investigate WORLDPAY credit card BIN failures | Improve 58.51% credit card approval | BIN-level analysis |
| P1 | Retrain model with processor-currency affinity | Prevent WORLDPAY+MXN toxic combinations | Feature engineering + retrain |
| P2 | Expand 3DS enrollment for high-value transactions | 3DS routes achieve 99-100% approval | Integration work |
| P2 | Document ELAVON Mar 27 configuration change | Understand +19pp spike for reproducibility | Investigation |
◍Key Findings
Critical observations from the past month's model performance. Apr 15 - Aug 14, 2026
1. WORLDPAY MXN Domestic Fallback Failure: When WORLDPAY receives overflow from MIT_BULK-primary MXN domestic strategies, approval rates collapse to 0-11%. This affects ~22K attempts/week and is the single largest source of failed transactions. The model must stop routing MXN domestic to WORLDPAY as fallback.
2. Y4 ADYEN ECOMM MXN: Only 15.89% approval rate on 491 attempts routed through WORLDPAY. The ADYEN integration via WORLDPAY is fundamentally broken for MXN.
3. 3DS Routes Near-Perfect: All 3DS-authenticated strategies show 99-100% approval rates regardless of processor. Expanding 3DS enrollment could significantly boost overall approval.
4. ELAVON Improvement: +4.07 pp week-over-week improvement, with a notable spike on Mar 27 (69% to 88%). Likely a configuration change. If intentional, this is a positive signal.
5. WORLDPAY Credit Card Gap: Credit card approval at 58.51% vs debit at 66.66%. This 8pp gap needs BIN-level investigation to identify problematic card issuers.
6. Retry Approval Rates Near Zero: Retries across all processors show 0.00%–5.46% approval (vs 90–96% on first attempts). WORLDPAY: 139,843 retries at 5.46%. MIT_BULK: 101,019 retries at 0.78%. ELAVON: 16,762 retries at 0.41%. AMEX: 3,332 retries at 0.00%. This ~90pp gap suggests retries are being sent to processors that will almost certainly decline again.
7. Massive Retry Volume Waste: 260,956 retry attempts across all processors with near-zero success. This represents significant wasted processing cost, potential card network penalties, and degraded user experience. Retry logic needs to consider decline reason codes and route to different processors.
◈Daily Approval Rate Trends
Day-by-day approval rate for each processor. Apr 15 - Aug 14, 2026 (122 days)
ELAVON spike: Approval rate jumped from 69% (Mon-Tue) to 88% (Thu-Sat), then settled at 82-83%. Investigate possible configuration change on March 27.
WORLDPAY uptrend: Gradual improvement from 70.28% to 74.58% over the week (+4.3 pp).
◉Processor Detail
Detailed breakdown of attempts, successes, errors, and approval rates. Apr 15 - Aug 14, 2026
| Processor | Total Attempts | Successes | Errors | Approval Rate | Volume Share | WoW Delta |
|---|---|---|---|---|---|---|
| MIT_BULK | 125,906 | 102,876 | 23,030 | 81.71% | 56.2% | -0.21 pp |
| WORLDPAY | 74,679 | 54,441 | 20,238 | 72.90% | 33.4% | -1.00 pp |
| AMEX | 15,476 | 14,147 | 1,329 | 91.41% | 6.9% | -0.35 pp |
| ELAVON | 7,862 | 6,251 | 1,611 | 79.51% | 3.5% | +4.07 pp |
◊Strategy Analysis
Top routing strategies by volume and their approval rates. Apr 15 - Aug 14, 2026
Top 15 Strategies (by volume) | Mar 8 - Apr 8
| Strategy | Total | Successes | Approval Rate |
|---|---|---|---|
| Y4 CREDIT WEB CEBUR MXN | 91,958 | 60,656 | 65.96% |
| Y4 ECOMM USD | 30,776 | 24,911 | 80.94% |
| Y4 NON-CREDIT WEB MXN | 29,295 | 17,309 | 59.09% |
| Y4 ATO/CCL MXN | 16,778 | 11,000 | 65.56% |
| Y4 CREDIT WEB CEBUR MXN - DIGITALAPI | 14,961 | 10,073 | 67.33% |
| Y4 WP ECOMM BAN-SAN MXN | 14,756 | 9,282 | 62.90% |
| Y4 AMEX WEB MXN | 9,998 | 9,367 | 93.69% |
| Y4 ATO/CCL USD | 7,028 | 6,151 | 87.52% |
| 3DS WP Y4 ECOMM USD X BIN | 5,146 | 5,143 | 99.94% |
| Y4 ECOMM USD ROW | 4,239 | 3,167 | 74.71% |
| N3 WP ECOMM USD | 2,356 | 1,926 | 81.75% |
| Y4 AMEX WEB USD | 2,240 | 2,083 | 92.99% |
| Y4 WP MOTO BAN-SAN MXN | 1,819 | 1,306 | 71.80% |
| Q6 ECOMM USD ROW | 1,191 | 915 | 76.83% |
| Y4 MOTO USD ROW | 1,064 | 791 | 74.34% |
Critical: The highest-volume strategy "Y4 CREDIT WEB CEBUR MXN" (92K attempts) runs at only 65.96%. Within this strategy, MIT_BULK achieves 76.23% but WORLDPAY fallback collapses to 8.68%.
Multi-Processor Split Routes (WORLDPAY fallback problem) | Mar 8 - Apr 8
| Strategy | MIT_BULK Rate | MIT_BULK Vol | WORLDPAY Rate | WP Vol | Delta |
|---|---|---|---|---|---|
| Y4 CREDIT WEB CEBUR MXN | 76.23% | 77,979 | 8.68% | 13,979 | -67.55 pp |
| Y4 NON-CREDIT WEB MXN | 69.78% | 23,941 | 11.24% | 5,354 | -58.54 pp |
| Y4 ATO/CCL MXN | 76.27% | 14,026 | 11.01% | 2,752 | -65.26 pp |
| Y4 CREDIT WEB CEBUR MXN - DIGITALAPI | 78.07% | 12,903 | 0.00% | 2,058 | -78.07 pp |
◇Decline Analysis
Hard decline vs soft decline breakdown by processor. Apr 15 - Aug 14, 2026
Decline Type Distribution | Mar 8 - Apr 8
Decline Breakdown Table | Mar 8 - Apr 8
| Processor | Approved | Soft Decline | Hard Decline |
|---|---|---|---|
| AMEX | 99.24% | 0.76% | 0.00% |
| MIT_BULK | 96.81% | 2.53% | 0.67% |
| ELAVON | 95.35% | 3.67% | 0.98% |
| WORLDPAY | 66.55% | 28.26% | 5.19% |
WORLDPAY soft decline rate (28.26%) is nearly double MIT_BULK's (15.96%). Soft declines are often retryable. WORLDPAY also has the highest hard decline rate at 5.19%.
○Card Type Breakdown
Approval rates segmented by card type (credit, debit, prepaid) per processor. Apr 15 - Aug 14, 2026
WORLDPAY credit cards: Only 58.51% approval rate vs 66.66% for debit -- an 8pp gap. ELAVON shows a similar pattern (62.88% credit vs 76.90% debit).
▶Retry Metrics
Retry vs first-attempt approval rates across processors and models. Apr 15 - Aug 14, 2026
Why retries matter: Retry attempts represent transactions that failed on the first processor and were re-routed. Comparing first-attempt vs retry approval rates reveals which processors recover well on retries and which models route retries most effectively.
First Attempt Rate
--
Across all processors
Retry Rate
--
Across all processors
Retry Recovery Lift
--
First attempt - retry delta
Retry Volume %
--
% of total attempts that are retries
First Attempt vs Retry — by Processor
Retry Volume Share — by Processor
Retry Breakdown by Processor
| Processor | 1st Attempt Vol | 1st Attempt Rate | Retry Vol | Retry Rate | Delta |
|---|---|---|---|---|---|
| AMEX | 331,467 | 92.73% | 2,554 | 0.00% | -92.73 pp |
| ELAVON | 131,905 | 88.35% | 4,637 | 29.54% | -58.81 pp |
| MIT_BULK | 2,457,386 | 79.63% | 56,443 | 5.87% | -73.76 pp |
| WORLDPAY | 1,871,209 | 78.56% | 66,205 | 2.30% | -76.26 pp |
Approval Rate by Attempt Number
| Processor | Attempt #1 | Attempt #2 | Attempt #3 | Attempt #4 | Attempt #5 |
|---|---|---|---|---|---|
| AMEX | 91.79% (254,502) | 89.81% (3,494) | 76.62% (77) | — | — |
| ELAVON | 90.06% (99,139) | 60.98% (5,225) | 37.35% (83) | — | — |
| MIT_BULK | 86.68% (1,762,569) | 39.48% (38,441) | 0.05% (13,022) | 0.00% (78) | 0.00% (4) |
| WORLDPAY | 30.46% (334,515) | 98.53% (906,349) | 87.51% (66,700) | 99.71% (2,763) | 100.00% (10,736) |
WORLDPAY retries actually improve: Attempt #1 at 68.54%, but attempts #2-3 jump to 81-82%, and #4-5 hit 99%+. This suggests WORLDPAY retry logic is effective — failed first attempts get recovered on subsequent tries. In contrast, ELAVON and MIT_BULK retries collapse to near 0% on attempt #2+.
Retry Performance by Model / Strategy
Retry Breakdown by Model × Processor
| Model / Strategy | Processor | 1st Attempt Vol | 1st Attempt Rate | Retry Vol | Retry Rate | Delta |
|---|---|---|---|---|---|---|
Note: applied_strategy contains routing rules, not Athia model names. Needs migration to athia_experiment_info:experiment_group for proper model attribution. | ||||||
Critical finding: Retry approval rates are near zero across all processors (0.00%–5.46%). First-attempt rates are 90–96%. This ~90pp gap means retries are almost never succeeding. Possible causes: (1) retries sent to same processor that declined, (2) hard declines being retried, (3) retry logic not considering decline reason codes.
WORLDPAY has the highest retry volume (139,843 attempts) with only 5.46% approval. MIT_BULK retries 101,019 attempts at 0.78%. These represent significant wasted processing cost and potential card network penalties.
◌Week-over-Week Comparison
Comparing this week (Mar 24-31) vs previous week (Mar 17-23). WoW snapshot as of Mar 31, 2026
| Processor | Prev Week Rate | Prev Week Vol | This Week Rate | This Week Vol | Delta |
|---|---|---|---|---|---|
| AMEX | 91.76% | 27,390 | 91.17% | 31,181 | -0.59 pp |
| ELAVON | 91.45% | 6,339 | 91.11% | 7,326 | -0.34 pp |
| MIT_BULK | 82.64% | 183,813 | 82.75% | 230,766 | +0.11 pp |
| WORLDPAY | 84.90% | 158,595 | 85.18% | 197,577 | +0.28 pp |
★Volaris Smart Router Analysis
Deep-dive into routing strategy performance, decline reasons, and error categories for Volaris. 30-day window
Note:
athia_experiment_info:experiment_group is not yet populated in transactions. Strategy names below are Deuna routing rules. Once experiment groups are enabled, this section will show per-model (TF Per-Processor vs XGBoost LTR) comparisons.
Total Volaris Attempts
4,970,692
30-day window
Top Strategy
100.0%
3DS WP Q6 ECOMM USD X BIN [WORLDPAY]
Worst Strategy
0.0%
Default [MIT_BULK]
Routing Rules Active
75
Distinct strategies
Top Routing Strategies — Performance by Processor
Strategy Performance Table
| Strategy | Processor | Volume | Approval Rate | Errors |
|---|---|---|---|---|
| Y4 CREDIT WEB CEBUR MXN | MIT_BULK | 1,294,954 | 79.38% | 266,976 |
| Y4 NON-CREDIT WEB MXN | MIT_BULK | 653,667 | 79.28% | 135,431 |
| Y4 ECOMM USD | WORLDPAY | 652,272 | 88.04% | 78,044 |
| Y4 WP ECOMM BAN-SAN-BBVA-BNMX MXN | WORLDPAY | 348,655 | 78.42% | 75,234 |
| Y4 AMEX WEB MXN | AMEX | 226,704 | 94.01% | 13,590 |
| Y4 ATO/CCL MXN | MIT_BULK | 200,679 | 77.29% | 45,573 |
| Y4 CREDIT WEB CEBUR MXN - DIGITALAPI | MIT_BULK | 182,673 | 80.98% | 34,736 |
| Y4 ECOMM USD ROW | WORLDPAY | 145,263 | 78.06% | 31,868 |
| RISKIFIED Y4 WEB USD | WORLDPAY | 116,059 | 89.90% | 11,724 |
| Y4 CREDIT WEB CEBUR MXN | WORLDPAY | 113,442 | 17.50% | 93,594 |
| Y4 ATO/CCL USD | ELAVON | 109,864 | 89.15% | 11,918 |
| — | MIT_BULK | 98,031 | 57.70% | 41,463 |
| Y4 WP ECOMM BAN-SAN MXN | WORLDPAY | 87,485 | 74.43% | 22,372 |
| 3DS WP Y4 ECOMM USD X BIN | WORLDPAY | 76,238 | 99.98% | 13 |
| Y4 AMEX WEB USD | AMEX | 63,158 | 94.18% | 3,677 |
| Y4 ADYEN BBVA CREDITO MXN (WEB) | WORLDPAY | 55,717 | 55.91% | 24,568 |
| N3 WP ECOMM USD | WORLDPAY | 52,876 | 85.57% | 7,629 |
| Y4 NON-CREDIT WEB MXN | WORLDPAY | 51,659 | 18.39% | 42,160 |
| — | WORLDPAY | 37,643 | 63.55% | 13,719 |
| RISKIFIED Y4 WEB MXN | MIT_BULK | 37,187 | 79.61% | 7,581 |
| Y4 WP MOTO BAN-SAN-BBVA-BNMX MXN | WORLDPAY | 27,765 | 75.59% | 6,778 |
| Y4 CREDIT WEB CEBUR MXN - DIGITALAPI | WORLDPAY | 24,540 | 9.05% | 22,319 |
| Y4 ATO/CCL MXN | WORLDPAY | 23,939 | 21.68% | 18,750 |
| Y4 WP ECOMM BAN-SAN-BBVA-BNMX MXN | MIT_BULK | 23,429 | 20.00% | 18,744 |
| Q6 ECOMM USD ROW | WORLDPAY | 22,984 | 81.59% | 4,232 |
Error Categories — by Processor
Decline Type Mix — by Processor
Error Category Breakdown
| Processor | Error Category | Count | % of Errors |
|---|---|---|---|
| AMEX | NA_no_identificado | 13,793 | 51.89% |
| AMEX | bank_rejected | 11,984 | 45.08% |
| AMEX | insufficient_amount | 519 | 1.95% |
| AMEX | deuna_settings | 93 | 0.35% |
| AMEX | user_settings | 90 | 0.34% |
| AMEX | deuna_setting | 52 | 0.20% |
| AMEX | max_attemps_quantity | 51 | 0.19% |
| ELAVON | bank_rejected | 6,764 | 33.16% |
| ELAVON | fraud_psp | 3,502 | 17.17% |
| ELAVON | insufficient_amount | 3,037 | 14.89% |
| ELAVON | user_settings | 2,092 | 10.25% |
| ELAVON | NA_no_identificado | 1,910 | 9.36% |
| ELAVON | fraud_bank | 930 | 4.56% |
| ELAVON | psp_int_settings | 875 | 4.29% |
| ELAVON | deuna_settings | 664 | 3.25% |
| ELAVON | max_attemps_quantity | 544 | 2.67% |
| ELAVON | psp_ext_settings | 82 | 0.40% |
| MIT_BULK | bank_rejected | 356,199 | 61.59% |
| MIT_BULK | insufficient_amount | 72,237 | 12.49% |
| MIT_BULK | user_settings | 60,415 | 10.45% |
| MIT_BULK | NA_no_identificado | 24,431 | 4.22% |
| MIT_BULK | deuna_settings | 21,689 | 3.75% |
| MIT_BULK | psp_int_settings | 21,625 | 3.74% |
| MIT_BULK | fraud_psp | 7,990 | 1.38% |
| MIT_BULK | fraud_bank | 6,903 | 1.19% |
| MIT_BULK | max_attemps_quantity | 4,693 | 0.81% |
| MIT_BULK | psp_ext_settings | 2,014 | 0.35% |
| MIT_BULK | unknown | 153 | 0.03% |
| WORLDPAY | bank_rejected | 277,894 | 57.24% |
| WORLDPAY | user_settings | 50,962 | 10.50% |
| WORLDPAY | insufficient_amount | 44,614 | 9.19% |
| WORLDPAY | NA_no_identificado | 23,898 | 4.92% |
| WORLDPAY | fraud_bank | 21,941 | 4.52% |
| WORLDPAY | psp_int_settings | 20,549 | 4.23% |
| WORLDPAY | fraud_psp | 20,163 | 4.15% |
| WORLDPAY | max_attemps_quantity | 14,504 | 2.99% |
| WORLDPAY | deuna_settings | 7,052 | 1.45% |
| WORLDPAY | psp_ext_settings | 3,586 | 0.74% |
| WORLDPAY | unknown | 187 | 0.04% |
| WORLDPAY | blacklist_psp | 117 | 0.02% |
Toxic Routing Combinations (Approval < 15%)
| Strategy | Processor | Volume | Approval Rate | Action |
|---|---|---|---|---|
| Y4 CREDIT WEB CEBUR MXN - DIGITALAPI | WORLDPAY | 24,540 | 9.05% | Block or reroute |
| Y4 WP ECOMM BAN-SAN MXN | MIT_BULK | 17,933 | 14.20% | Block or reroute |
| Y4 ADYEN WEB BBVA DEBITO MXN | WORLDPAY | 4,554 | 5.88% | Block or reroute |
| Y4 ADYEN ECOMM BBVA WEB MXN | WORLDPAY | 1,829 | 1.09% | Block or reroute |
| Y4 ADYEN WEB BBVA DEBITO MXN | MIT_BULK | 1,690 | 7.10% | Block or reroute |
| Default | MIT_BULK | 1,566 | 0.00% | Block or reroute |
| Y4 ECOMM GBP ROW | WORLDPAY | 1,020 | 12.16% | Block or reroute |
| Y4 ADYEN ECOMM BBVA WEB MXN | MIT_BULK | 929 | 6.46% | Block or reroute |
| Y4 ADYEN BBVA CREDITO MXN | WORLDPAY | 630 | 2.86% | Block or reroute |
Toxic combinations are strategy+processor pairs with <15% approval on significant volume. These represent wasted attempts that should be blocked by the model or routing rules.
⚖Sampling Bias Audit
Volaris Multi-Horizon Forensic Audit — DNN vs Heuristic population distinguishability across 3 rolling windows (24h / 7d / 15d). TVD < 2% = indistinguishable; 2-5% = minor drift; ≥5% = divergent.
24h Verdict
—
—
7d Verdict
—
—
15d Verdict
—
—
15d 6D avg TVD
—
—
TVD per Feature × Window
Total Variation Distance between control (Heuristic) and treatment (DNN) for each pre-routing feature. Lower = more similar populations.
Group Sizes per Window
| Window | Control (n) | Treatment (n) | Total | Split % | Balance |
|---|---|---|---|---|---|
| No data — run refresh_metrics.py to populate | |||||
TVD per Feature (7d window)
| Feature | TVD | Status |
|---|---|---|
| No data — run refresh_metrics.py to populate | ||
⚙Model Reference collapsed
All ML models and routing strategies currently deployed in production.
Production Model Types
| Model | Framework | Versions | Purpose |
|---|---|---|---|
| processor_selector | XGBoost LTR | v1, v2, v2_2, v3, v4 | Rank and select optimal payment processor for each transaction |
| volaris_smart_router | TensorFlow SavedModel | v1 (per-processor) | Per-processor approval probability prediction (140-feature encoder) |
| retry_predictor | XGBoost LTR | v1, v2, v2_1, v2_2 | Predict whether a failed transaction should be retried |
| installment_optimizer | scikit-learn RF | — | Predict optimal installment options for customer |
Routing Strategies
| Strategy | Model Used | Description | Status |
|---|---|---|---|
| tf_per_processor | volaris_smart_router | Scores each processor independently with TF model, then ranks. Default for Volaris. | Default |
| xgboost_ltr | processor_selector | Scores all processors in a single XGBoost batch using Learning-to-Rank. | Active |
Volaris Per-Processor TF Models
| Processor | Model Format | Features | Encoder |
|---|---|---|---|
| worldpay | TF SavedModel | 140 (13 z-score + 127 OHE) | volaris_tf_encoder.py |
| amex | TF SavedModel | 140 (13 z-score + 127 OHE) | volaris_tf_encoder.py |
| elavon | TF SavedModel | 140 (13 z-score + 127 OHE) | volaris_tf_encoder.py |
| mit_bulk | TF SavedModel | 140 (13 z-score + 127 OHE) | volaris_tf_encoder.py |
| uatp | TF SavedModel | 140 (13 z-score + 127 OHE) | volaris_tf_encoder.py |
| bbva | TF SavedModel | 140 (13 z-score + 127 OHE) | volaris_tf_encoder.py |
Infrastructure
| Component | Detail |
|---|---|
| Serving | athia-model-server sidecar (Python FastAPI) alongside Go API |
| Model Storage | S3 → EFS mount at /mnt/models |
| Training Repo | DATA-Athena-Snowflake (Python, ClearML tracking) |
| Serving Repo | athena-platform (Go + Python sidecar) |
| Infra | Spacelift — terraform-athia |
| Dashboard | Athia Experiments Dashboard |