AI agent and digital experience analytics from event streams to outcomes
Conviva analyzes billions of digital interactions—video streams, web apps, and AI agents—to surface behavioral patterns and business outcomes. The tech stack (PyTorch, TensorFlow, Spark, Flink, Kafka, Druid, ClickHouse) reveals a company built for high-volume event processing and real-time analytics; active projects on agent experience analytics and next-gen streaming systems signal a pivot from video QoE toward AI observability, matching hiring velocity that leans heavily engineering and product.
Conviva operates a telemetry and analytics platform that ingests full-census client-side data from digital properties—originally video streaming, now expanding to web applications and AI agent interactions. The platform uses a stateful analytics engine to process high-volume event streams, identify behavioral patterns, and connect them to business outcomes in real time. Customers include product, marketing, and engineering teams at mid-market and enterprise companies. The company is based in Foster City, CA, founded in 2006, and currently operates across the United States, China, and India.
Core stack includes PyTorch, TensorFlow, Apache Spark, Flink, Kafka, Druid, and ClickHouse. Supporting tools: Python, Scala, Java, Rust, Haskell, Erlang, and C++. Infrastructure: Hadoop, HDFS, Iceberg, Hudi. Operational: Jira, Confluence, Slack, Microsoft 365.
Foster City, California. The company also actively hires in China and India, alongside United States-based roles.
Other companies in the same industry, closest in size