Our open source framework for modeling streaming relational data from many heterogeneous sources and for querying it using Datalog and GraphQL.
Some modeling trade-offs you make with Kafka will be frozen in time. kplex eases the pain and reduces inertia, should these decisions come back to bite you.
Data InfrastructureWe plan and implement evolvable data storage solutions and service runtimes on bare metal, hybrid, or cloud deployments. We offer guidance throughout the initial setup and configuration and provide ongoing maintenance and DevOps training.
- + Simple, maintainable, evolvable as you grow
- + Consistent environments across development and all stages of deployment
- + Designed for security, efficiency, and compliance with privacy regulations
Data SolutionsOur software platform 3DF serves as a scalable foundation for combining many diverse data streams and delivering real-time results to interactive queries. We will customize it to your needs — with powerful algorithms and predictive models.
- + Real-time answers to complex questions across all your data
- + Automatic enforcement of access policies and business-critical compliance guarantees (GDPR)
- + Consistent interpretation and sub-second propagation of results throughout your organization
Web ApplicationsWe create fully customized, intuitive interfaces, exploratory data visualizations, and collaboration tools for the web and mobile platforms. We strive for technology that is both calm and powerful.
- + Reflect the state of your business in real-time
- + Sophisticated, exploratory interactions and support for data-driven decision making
- + Hassle-free collaboration with full support for offline use
About Time: An Introduction to Timely Dataflow [talk]
This is a conference talk Malte and Niko gave at Data Council Barcelona 2019.
kplex - Decoupling Partitioning in Kafka
We discuss partitioning — arguably the highest impact decision in any Kafka setup — and introduce kplex, a tool for repartitioning Kafka topics consistently and on-the-fly, allowing you to unlock concurrency and perform correct stateful processing across partitions.
Incremental Functional Aggregate Queries
This blog post concludes a research project I’ve done together with Frank McSherry earlier this year at ETH Zurich.