How Factorial builds real-time data products with Confluent and Tinybird

Founded in 2016, Factorial expanded from basic HR software into an integrated HR, finance, and people-operations platform serving more than 8,000 businesses across 65 countries, but its original AWS-based batch data lakehouse could not provide the fresh data or low-latency queries needed for user-facing features. Scheduled MySQL snapshots, Spark processing, S3 storage, and Athena queries supported internal reporting but left data hours or days old and query responses ranging from seconds to minutes. To avoid separate developer-built systems and the operational burden of self-managed infrastructure, Factorial adopted Confluent Cloud and Tinybird, using MySQL change data capture through Debezium to stream changes into Kafka and Tinybird to enrich real-time streams with historical data and expose them through low-latency APIs. The managed architecture reduced data freshness to seconds and average production query times to under 50 milliseconds, while allowing a two-person data team to launch its first production feature within a month and more than 12 real-time user-facing features over the following six months.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.