Rerun

Hugging Face's LeRobot documentation integrates Rerun into its dataset-visualization command, opening camera streams, robot states and actions for a selected local or Hub dataset episode.

Rerun provides an open-source and commercial data layer for Physical AI, covering logging, storage, SQL/dataframe querying, visualization, transformation and training workflows for multimodal robotics data.

Rerun provides an open-source and commercial data layer for Physical AI, covering logging, storage, SQL/dataframe querying, visualization, transformation and training workflows for multimodal robotics data.

Rerun documents DataFusion-based streaming dataset queries and a sparse event-layer technique that permits catalog-server filter pushdown to avoid scanning all recording rows; examples show registration of the layer and filtering by non-null markers.

Rerun is building a unified columnar data architecture and developer toolchain for multi-rate, multimodal robotics data from raw recording through querying, visualization and training.

Rerun is building a unified columnar data architecture and developer toolchain for multi-rate, multimodal robotics data from raw recording through querying, visualization and training.

Rerun announced $17 million in seed financing led by Point Nine with Costanoa, Sunflower Capital and Seedcamp participating.

Rerun describes a distributed remote-first organization with founders in Stockholm and quarterly in-person gatherings; its careers page lists backend, infrastructure, forward deployment, strategic operations, talent and community functions.

Rerun Hub provides a production backend with cataloging, byte-range indexing, storage access and streaming over object stores, but public evidence of very large enterprise deployment scale remains limited.

Rerun documents logging user-defined data serialized as Apache Arrow without recompilation, with Python/Rust AnyValues helpers and custom component protocols. This concrete data-format interoperability supports inferred portable multimodal data engineering skills.

Rerun work spans robotics data infrastructure, Rust/C++/Python, multimodal data, visualization, storage, query engines and training pipelines, creating highly portable physical-AI infrastructure skills.

Rerun work spans robotics data infrastructure, Rust/C++/Python, multimodal data, visualization, storage, query engines and training pipelines, creating highly portable physical-AI infrastructure skills.