ArcticDB

Health v2 · maintenance only

Flags: no_license

Adoption not part of the score

2493 stars · 214 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

ArcticDB is a high-performance DataFrame database for time series and tick data, developed at Man Group for quantitative data science. It stores and retrieves Pandas DataFrames natively via a Python API, sitting directly on commodity object storage (S3, Azure Blob, GCP, LMDB) with no servers to provision.

Use cases

• store and query billions of rows of time series data from Python
• persist pandas DataFrames with versioning and point-in-time queries
• build a tick data store for market data in quantitative finance
• backtest trading strategies against large historical datasets
• store evolving-schema dataframe data on S3 without managing a database server
• handle wide cross-sectional data with hundreds of thousands of columns

When to choose

• you work in Python with pandas/polars and need fast dataframe storage at scale
• you want a serverless datastore on top of S3/Azure/GCP object storage
• you need immutable versioning and time-travel queries over time series
• you are doing quantitative research, backtesting, or market data storage

When to avoid

• you need SQL queries or joins across relational tables
• you need a general-purpose OLTP database with transactions
• your stack is not Python-centric
• you need real-time streaming ingestion with sub-millisecond latency

Facets

library · maturity active

database data-science etl databases data-science fintech time-series python windows dataframe-database timeseries tick-data object-storage pandas quantitative-finance serverless-database versioned-data linux macos

Member repositories

For agents

markdown · JSON · MCP: product_card(name="man-group/ArcticDB")