Home AIS AIOS Assistant
Solutions🔧 Contractor Sites⚡ All Services
Case Studies Blog About
Book a Call

Case studies — modeled outcomes

Below: three representative deployments and what Nova would unlock at their scale. These are modeled on real PH market data — not signed customer announcements.

Case 01 · Collections AI

Top-5 PH commercial lender

~₱29.9B outstanding · 4.7% default rate · ~100 collections agents

Problem

Recovery rate stuck near 20% on defaulted accounts. RPC (right-party-contact) at ~3% on cold dialer. Compliance review takes weeks per quarter. Settlement decisions made by tenure not data.

What we'd plug in

Nova Finance sits on top of the existing LMS at the missed-payment branch. Early-warning model scores every loan 30-90 days out. Risk segmentation routes Tier-1 outreach to AI voice + SMS agents. Compliance guard logs every contact for FDCPA / RA 11765. Settlement optimizer recommends offer % per account.

Modeled outcome (12 months)

Recovery rate moves 20% → 30% on the defaulted pool. Headcount drops 40% on Tier-1 outreach. RPC lifts ~4× via best-time + best-channel routing. Legal-filter cuts unrecoverable-account fees ~35%.

Case 02 · MFI / Cooperative

Mid-tier microfinance cooperative

~₱1.2B outstanding · 8.2% default rate · ~15 field officers

Problem

Group-lending model means physical visits eat the margin on small loans. Officers spend 60% of time driving between barangays. Defaults are visible only after they happen.

What we'd plug in

Nova Finance + AI Voice agent for SMS-first outreach in Tagalog + Cebuano. Officers get a prioritized route each morning — visit only the accounts where in-person actually changes the outcome. Group-leader notifications via Messenger.

Modeled outcome

Field-visit volume cut in half. Recovery rate on missed-payment branch lifts from ~25% to ~38%. Officers cover 1.8× more accounts. Compliance with BSP MFI Circular 1011 logged automatically.

Case 03 · BNPL Operator

BNPL operator with ~10k active loans

~₱180M monthly issuance · 6.1% delinquency · 24-agent collections team

Problem

BNPL volume scaled faster than the collections team. New cohort delinquencies hit before the team can call. Per-account economics are tight — every minute of agent time matters.

What we'd plug in

Pre-default early warning. AI voice + SMS agent handles 70% of Tier-1 outreach autonomously. Human agents only touch accounts where the AI flagged "human required." Settlement optimizer maximizes recovery × acceptance probability.

Modeled outcome

Collections headcount needs cut 38% at current volume. Recovery rate on delinquencies moves 22% → 35%. Time-to-first-contact post-default drops from 3 days to ~4 hours.

Case 04 · Contractor AI + Website

3-person HVAC crew — no website, 40% missed calls

$35K/mo revenue · 8 Google reviews · no website · no after-hours coverage

Challenge

Owner-operator losing ~$13,650/mo in missed calls while on the roof. No online presence. Competitors had 85+ reviews. Homeowners searching "HVAC repair Phoenix" never found him.

Nova Stack Deployed

Growth tier ($1,297 + $199/mo): 5-page contractor website with job-photo gallery, service area pages for 5 ZIP codes, AI phone agent, missed-call text-back, review automation, AI chat widget, CRM pipeline, follow-up engine, reactivation bot. Live in 7 days.

Result

+$14,200/mo recovered revenue. Google ranking: page 3 → local 3-pack. Reviews: 8 → 47 in 60 days. Crew utilization: 60% → 92%. Now hiring a 4th tech. Read the full case study →

Want this modeled on your actual numbers?

Send us your outstanding portfolio + default rate. We'll model your additional recovery, agent efficiency, and payback in a custom 1-page memo within 24h.

Run the ROI calculator → Or book a call →