AI-powered identity verification platform for enterprise onboarding and fraud prevention
Incode builds an identity verification platform using Python, PyTorch, and mobile SDKs (iOS/Android) to detect fraud and authenticate users at scale. The tech stack reveals a ML-first architecture—PyTorch for model development paired with Core ML and AVFoundation for on-device inference—while active projects in deepfake detection, facial recognition, and liveness modeling show heavy investment in biometric AI. The hiring mix (13 engineers, 9 data, 9 sales) and migration from Java to Kotlin signal both platform modernization and sales-driven growth in a compliance-heavy market.
Incode is an AI-powered identity verification platform serving enterprises that need to verify users, prevent fraud, and stay compliant during digital onboarding and authentication. Founded in 2015, the company operates from San Francisco with 501–1,000 employees spread across the US, Serbia, Mexico, Colombia, and Peru. The platform combines deterministic verification, deepfake detection, and intelligent orchestration to help customers distinguish legitimate users from fraudsters across KYC, AML, and identity management workflows. Incode was named a Leader in the 2025 Gartner Magic Quadrant for Identity Verification for a second consecutive year.
Incode runs on AWS with Kubernetes, Terraform, and Apache Kafka. Core services use Python and SQL (Redshift), while mobile SDKs span iOS (Swift, Objective-C, Core ML) and Android (Kotlin, RxJava). Monitoring uses Prometheus and Loki; ML workloads run on PyTorch.
Active projects include iOS SDK development and Kotlin migration, cloud-native infrastructure scaling, document processing and facial recognition models, liveness detection, and identity ecosystem expansion. Pain points center on scaling global infrastructure, reducing fraud, and ensuring PII compliance.
Other companies in the same industry, closest in size