Ad network for AI chat applications connecting publishers and advertisers
Koah operates a two-sided marketplace for GenAI applications, handling real-time bidding, conversion prediction, and latency-critical matching. The stack—Kafka, ClickHouse, PostgreSQL, Debezium—reveals a data-heavy matching engine with streaming ingest and OLAP analytics. The engineering-dominant hiring profile (6 of 7 roles) and project focus on matching algorithms, p99 latency optimization, and conversion regression models indicate the company is building infrastructure-grade ad-serving logic rather than early-stage experimentation.
Koah is a GenAI ad network enabling publishers inside AI chat applications to monetize without degrading user experience, while providing advertisers direct access to high-intent users within generative AI environments. Founded in 2024 and based in San Francisco, the company operates a two-sided platform across web and mobile (React, React Native, Flutter, iOS, Android). Core challenges include balancing publisher and advertiser incentives, managing p99 latencies in critical paths, and defining an economic model suited to AI-native monetization rather than traditional ad unit structures.
Terraform, AWS, PostgreSQL, ClickHouse, Kafka, Debezium, Redis, Python, Ruby on Rails, React, TypeScript, Grafana, Loki, Tempo, Mimir, Cloudflare, Tailscale, and iOS/Android/Flutter SDKs.
Real-time bidding and matching algorithms, conversion prediction regression models, p99 latency reduction in write paths, privacy-preserving advertiser clustering, and data pipelines streaming PostgreSQL changes to ClickHouse for analytics.
Other companies in the same industry, closest in size