Aarki operates a programmatic DSP built on four owned data centers processing 6M+ queries per second, with a tech stack spanning TensorFlow, PyTorch, Kafka, and ClickHouse. The hiring mix reflects dual scaling: 7 marketing roles and 5 sales roles alongside 5 engineering and 3 data positions, suggesting a push to expand customer acquisition and revenue operations. Active projects around low-latency prediction infrastructure and fraud detection indicate the company is hardening core campaign-optimization systems while chasing forecasting accuracy and user-response prediction.
Notable leadership hires: Head of Sales
Aarki (operating as RZR) builds a performance-marketing platform that combines user acquisition, retargeting, connected TV, and influencer campaigns through a unified DSP backed by proprietary machine-learning models. The platform routes decisions through owned infrastructure: four data centers handle 6M+ queries per second, feeding real-time optimization across creative variants, bidding, and audience segmentation. The company serves performance-driven brands at scale, with offices in seven countries (San Francisco, New York, London, Bangalore, Beijing, Manila, Seoul) and a workforce of 51–200. Revenue model centers on managed services and platform adoption fees tied to campaign performance.
Core: TensorFlow, PyTorch, Kafka, ClickHouse, Apache Spark, Flink. Data pipeline: HDFS, Fluentd, SQL. Visualization: Tableau, Sisense. Infrastructure runs on C++, Java, Rust, Go. Creative tools include Adobe suite and Figma.
Low-latency prediction infrastructure, fraud detection, programmatic DSP data infrastructure, campaign-performance optimization, and multi-scenario forecasting. Current pain points include churn prevention, forecasting accuracy, and bid-landscape prediction.
Other companies in the same industry, closest in size