Protecting people from AI powered scams
AI that screens calls and messages before they reach you.
Available soon for iOS and Android
The Problem
AI-powered fraud is scaling faster than any defense.
Voice cloning, synthetic identity, and LLM-generated scripts have turned the telephone into the most dangerous attack vector in the world.
$4.9B
Lost by adults 60+ to cybercrime in 2024
43% YoY increase
1,200%
Increase in AI-powered scam calls in 2025
289
Scam call attempts per person per year in Argentina
$40B+
Projected global fraud losses by 2027
“Voice cloning takes 3 seconds of audio. 1 in 4 Americans cannot distinguish a cloned voice from the real person.”
Vision & Architecture
The Infrastructure of Digital Trust.
We build the AI pipelines and real-time security protocols to protect human communication from synthetic deception. Strict TypeScript architecture and edge-optimized inference — designed to be deployed at telecom scale.
Module 01
High-Scale Inference
Real-time analysis of voice and metadata at the network edge. Sub-100ms latency for fully transparent filtering.
Module 02
Synthetic Detection
Proprietary models trained to distinguish human intent from generative AI fraud — identifying linguistic patterns unique to LLM-driven attacks.
Module 03
Zero-Trust Networking
A secure communication perimeter that processes all sensitive data locally, ensuring privacy-by-design at the protocol level.
1.2B+
Threats Indexed
2ms
Sync Latency
Global Intelligence
Global Blacklisted Lines Database.
A real-time, decentralized intelligence layer that identifies and blacklists fraudulent numbers across the globe — instantly. Centelo tracks threats across borders to prevent attacks before they even start.
Reinforcement Learning from Every Call
Centelo's AI doesn't just block — it learns. Every interaction reinforces our detection engine, creating a threat model that evolves faster than the attacks it defends against.
Market Opportunity
Three markets. One compounding data moat.
Our go-to-market is consumer — but our technology is infrastructure. Every market we enter generates proprietary threat data that compounds into a moat no competitor can buy.
Launch Market
Argentina
289 scam attempts per person per year. Centelo's first production deployment — proving real-world efficacy at scale.
Expansion
Brazil
$54B in fraud losses in 2024. Massive underserved opportunity with no dominant player in real-time call protection.
Scale
United States
$4.9B in elderly losses. The US Latino diaspora is our natural bridge market into the largest telecom spend in the world.
Enterprise & Telco Layer
We sell the intelligence. Telcos and enterprises deploy it.
Beyond consumer protection, Centelo licenses its global blocklist database, fraud vector signals, and real-time AI screening APIs to carriers, insurance companies, and enterprise platforms. Your network protected at the infrastructure level — before a single call reaches your customers.
Telcos
Real-time fraud signal APIs for carrier-level call filtering.
Insurance & Fintech
Identity verification and synthetic voice fraud detection at scale.
Elder Care Platforms
White-label protection for care networks serving vulnerable populations.
Revenue streams: Direct-to-consumer ($3.99–$7.99/mo) · B2B API licensing to Telcos & Enterprise · Threat intelligence data subscriptions
Team
The Team
Matias Paterlini
Founder & CEO
Building Centelo AI to protect the families that matter most.
Buenos Aires · San Francisco
LinkedIn →Partner With Us
Building the Future of Digital Safety.
We are seeking investors and partners who share our vision of a scam-free world. For investment inquiries, reach out directly.
arrow_forwardhello@centelo.ai