Neuralake - Complex Data, Simple Systems
OVERVIEW
EXPERIENCE | In Person |
---|---|
TYPE | Breakout |
TRACK | Data Engineering and Streaming |
TECHNOLOGIES | Delta Lake, Developer Experience, ETL |
SKILL LEVEL | Intermediate |
DURATION | 40 min |
DOWNLOAD SESSION SLIDES |
Enabling telepathy and vision restoration is no simple task, and Neuralink’s data reflects that. Neuralink manages a tightly-woven, sprawling web of information, connecting neural signals, manufacturing telemetry, in vivo study results, and more. Neuralink’s data is highly multimodal, including billion-pixel microscopy images, terabytes of raw brain data, and micron-scale surgical videos. In order to develop easy access to such complex data, Neuralink has developed Neuralake, a multimodal data lake that allows users to easily query and retrieve data. Neuralake developed a modern Rust based data stack, including polars, Delta Lake, Apache Arrow, and Apache Datafusion. This technology allows for simple, elegant systems that allow for easy data ingestion and access at scale. Neuralink maintains real-time data ingestion systems, federated multi-datastore querying, and fast data retrieval. Without the overhead of JVM or complex distributed systems, the Neuralake stack scales from a single developer laptop to a stateless distributed cluster
SESSION SPEAKERS
Gautham Acharya
/Software Engineering Lead
Neuralink