twosigma/flint

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1177 stars · 200 forks observed · 2026-08-28

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

Flint is Two Sigma's open-source time series library for Apache Spark, built around a time series aware TimeSeriesRDD data structure. It provides highly optimized, parallel time series operations that exploit temporal ordering for locality-based optimizations on large-scale and high-frequency data.

Use cases

• analyze large-scale time series data in Spark
• compute rolling window aggregations over high frequency financial data
• run panel data analysis across many instruments
• resample and summarize tick data at scale
• join and align multiple time series by timestamp
• process IoT sensor time series with Spark

When to choose

• you already use Spark 2.x and need efficient time series operations
• you work with high-frequency or panel data too large for a single machine
• you want locality-aware optimizations that respect temporal ordering

When to avoid

• you use Spark 3.x, which Flint does not support
• your time series data fits in pandas or polars
• you need an actively developed library with recent releases

Facets

library · maturity maintenance

data-science etl analytics time-series big-data data-science fintech jvm python apache-spark timeseries distributed-computing panel-data

Member repositories

For agents

markdown · JSON · MCP: product_card(name="twosigma/flint")