Programmatic DSP for mobile app user acquisition and retargeting with ML-driven targeting
Bigabid operates a demand-side platform (DSP) for mobile app marketers, built on Python, Spark, Flink, and PyTorch—a data-engineering-first stack optimized for real-time bidding and ML inference at scale. Active projects span feature stores, vector databases, and GPU cluster scaling; pain points cluster heavily around ML infrastructure (observability, GPU scaling, latency constraints), cloud cost, and revenue expansion—a pattern that suggests the company is maturing from core DSP capabilities into a multi-tenant, infrastructure-heavy platform serving high-budget app clients.
Bigabid is a DSP (demand-side platform) that helps mobile app marketers acquire and retarget users through programmatic advertising. The platform uses machine learning to analyze real-time behavioral and contextual data, targeting high-lifetime-value users across 1000+ app categories. Founded in 2016 and headquartered in Tel Aviv, the company employs 51–200 people with engineering and data science as primary functions. They work with app publishers and advertisers across iOS and Android ecosystems, positioning themselves on model transparency and post-IDFA audience precision.
Bigabid's stack centers on Python, SQL, Apache Spark, Apache Flink, Ray, and Dask for data processing; PyTorch and TensorFlow for ML; AWS for cloud infrastructure; Kubernetes for orchestration; and Datadog, Grafana, and Prometheus for observability.
Recent projects include feature store development, vector databases, GPU cluster scaling, deployment pipeline enhancement, cloud infrastructure evolution, enterprise-grade real-time data stores, and revenue expansion through multi-million-dollar account growth and ML technology upsell.
Other companies in the same industry, closest in size