AI-native risk decisioning platform for fraud, credit, and compliance
Oscilar operates a no-code risk platform built on a polyglot stack (Java, Python, Go, Kafka, ClickHouse, Spark, Dask) designed for real-time decisioning at scale. The engineering-heavy org (22 engineers, 5 data specialists) is actively shipping a no-code configuration layer, advanced fraud models, and observability infrastructure — suggesting they're moving beyond single-use detection toward a unified decisioning fabric. Pain points cluster around latency, resilience under traffic spikes, and multi-region reliability, which the architecture (Kafka + ClickHouse + Spark) is built to address.
Oscilar provides a unified risk decisioning platform that consolidates fraud prevention, credit underwriting, onboarding risk assessment, and AML compliance into a single interface. The product uses agentic AI and anomaly detection to automate high-stakes decisions in real time, targeting financial services and fintech companies. Founded in 2021 and headquartered in Palo Alto, the company is built by engineers who previously owned risk systems at Google, Meta, Uber, Citi, and J.P. Morgan. The 51–200-person team is actively hiring across the US, Poland, and Brazil, with senior and staff engineers leading core platform work.
Java, Python, Go, Kafka, ClickHouse, Apache Spark, Dask, Kubernetes, Docker, PostgreSQL, Redis, DynamoDB on AWS, GCP, and Azure. Jenkins, Terraform, and Pulumi for CI/CD and infrastructure.
No-code risk configuration platform, advanced fraud detection models, multi-region platform resilience, device fingerprinting SDKs, credit bureau integrations, observability tooling, and CI/CD pipeline automation.
Other companies in the same industry, closest in size