Pricing models and real-time data infrastructure for sports betting and trading
Mustard Systems builds pricing models and data infrastructure for licensed bookmakers and professional traders. The stack—Snowflake, Kafka, dbt, Python, PostgreSQL—reflects a data-first architecture optimized for real-time feature generation, but the active pain points (data warehouse cost optimization, low-latency feed reliability, pipeline freshness observability) suggest they're hitting scaling limits common to young trading-tech companies. Projects spanning web scraping, computer vision on cricket broadcasts, and liquidity provider reverse-engineering indicate they're building proprietary data collection to solve liquidity and signal sourcing constraints.
Mustard Systems develops pricing models and operates data infrastructure for sports betting operators and professional gamblers. Founded in 2014 as a spin-out from a quantitative research firm, the company is based in London with 11–50 employees. The business model centers on selling pricing services and modeling capabilities to licensed bookmakers, augmented by internal data pipelines that ingest market feeds, historical sports footage, and liquidity data. The team is weighted toward data engineering (5 roles) and software engineering (3 roles), with mid and senior-level hiring dominating recent activity.
Mustard Systems uses Snowflake, PostgreSQL, Kafka, dbt, Python, JavaScript, TypeScript, Go, Redis, RabbitMQ, and monitoring tools including OpenTelemetry, Prometheus, and Grafana. All infrastructure runs on AWS and Linux.
Projects include improving trading systems, scaling data pipelines, developing low-latency real-time feeds, expanding liquidity provider access through web scraping, and building computer vision systems to extract data from cricket broadcasts for sports analytics.
Other companies in the same industry, closest in size