ArcticDB is a server‑less, high‑performance DataFrame database for quantitative data science. It stores Pandas or Polars DataFrames directly on object storage, delivering instant, versioned queries on billions of rows and hundreds of thousands of columns in seconds, all through a native Python API that integrates with common data‑science tools.
Funding
Funding not disclosed
Founders
Product
Problem
Quantitative analysts often struggle with processing and querying massive time‑series datasets because traditional databases require extensive data transformation, server provisioning, and cannot natively handle DataFrames at petabyte scale. This limits productivity when building, testing, and deploying trading models that need billions of rows and hundreds of thousands of columns.
Solution
ArcticDB is a high‑performance, serverless DataFrame database designed for quantitative data science. It stores and retrieves data directly as Pandas (or Polars) DataFrames through a native Python API, eliminating the need for data format conversion. Built on commodity object storage, the library scales from zero to unlimited users without managing servers or clusters. Its immutable, versioned storage lets users query data as it existed at any point in time, while supporting both dynamic and static schemas. The query engine delivers cross‑sectional views and time‑series analytics on billions of rows in seconds, enabling quants to focus on model development rather than data engineering.
Target Audience
Primary users are quantitative analysts, data scientists, and developers at hedge funds, asset managers, investment banks, and other firms that require high‑speed, large‑scale time‑series data processing within Python environments.
Features
- Processes billions of rows and hundreds of thousands of columns in seconds with an instantly queryable DataFrame format
- Serverless architecture that runs directly on object storage (S3, Azure Blob, GCP) with no dedicated servers or clusters
- Schemaless design with optional static schemas and built‑in versioning for immutable, point‑in‑time queries
- Native Python library (pip install arcticdb) with full Pandas and Polars integration and compatibility with common data‑science tools
- Flexible cross‑sectional and time‑series query capabilities optimized for quantitative workloads
- Seamless integration into Jupyter, CI pipelines, and existing Python analytics stacks