PhyloAI connects transaction feeds, settlement records, support tickets, MCA/GST filings and merchant CRM data into one ontology — delivering churn risk alerts, anomaly detection, revenue leakage flags and retention actions inside your PSP platform.
Relationship managers rely on merchant self-reports. PhyloAI reads what merchants don't flag — volume deceleration, settlement anomalies, support escalations and competitor PSP onboarding that precede account closure.
By the time volume drops appear in monthly reports, merchants have often already onboarded with a competitor PSP. Early signals live in daily transaction and settlement data.
Authorisation spikes, refund clusters and settlement mismatches are invisible at scale. Without ontology-driven rules, anomalies are found only during fraud investigations.
MDR waivers, promotional pricing and custom fee schedules create silent revenue gaps. Exceptions approved in CRM rarely match what settlement actually collects.
Most connectors are custom-built. PhyloAI maps your merchant ontology first, then connects the sensors that feed it.
Authorisation volume, ticket size, refund rates and decline patterns — monitored per merchant and segment.
Settlement timing, reconciliation gaps and MDR collection mismatches surfaced at T+1, not month-end.
Escalation velocity, chargeback disputes and onboarding friction — early churn indicators before volume drops.
Regulatory filings, business status changes and compliance gaps that signal merchant distress or pivot.
PhyloAI builds a merchant-specific ontology: which KPIs matter per segment, what thresholds trigger churn alerts, how signals combine into risk scores, and who owns each retention action.
MID, segment, MDR tier, tenure
Volume, ticket size, refunds, settlement
Thresholds, combinations, weights
Converging signal confidence
RM outreach, fee review, escalation
Ask what happens if at-risk merchants continue current trajectory — or if you intervene now. Scenarios grounded in live transaction and settlement data, not static cohort spreadsheets.
3 merchants projected to churn within 6 weeks. Combined revenue loss: ₹18.6L/month based on current GMV decline rate.
Churn probability drops 40–55% if retention pricing offered within 7 days. Recovers an estimated ₹11.2L/month in at-risk GMV.
2 of 3 merchants show dual-routing patterns. Recommend immediate executive escalation and custom fee proposal.
Transaction and settlement signals monitored continuously
Ontology explains why signals matter per merchant
Churn scenarios model retention outcomes
Retention actions assigned to relationship owners
Outcomes refine future churn rules
Continuous health scores per merchant — volume trends, settlement reliability, support friction and CRM engagement woven into one view.
See health dashboard →Authorisation spikes, refund clusters and settlement mismatches flagged with merchant context — ranked by severity and revenue impact in ₹.
See anomaly examples →MDR waivers, fee exceptions and promotional pricing gaps detected at settlement. Reconciliation mismatches surfaced before month-end close.
See leakage alerts →Churn alerts, merchant Q&A and retention recommendations embedded directly inside your PSP platform — no separate BI tool required.
See embedded view →Every retention outcome feeds back into your merchant ontology — churn models get sharper with each intervention cycle.
Relationship manager reaches out to 3 at-risk merchants based on volume decline alerts. Retention offer logged in PhyloAI.
Did volume stabilise? Did the merchant stay? Was the churn prediction accurate? Outcome data captured automatically from transaction feeds.
Signal weights, volume thresholds and combination rules updated based on which interventions actually prevented churn vs. false positives.
Each cycle makes churn scoring more accurate for your specific merchant segments, MDR tiers and competitive landscape.
Tell us about your merchant book. We'll map your ontology and show exactly which signals PhyloAI would monitor across your PSP platform.