finds.dev← search

// the find

unytics/bigfunctions

★ 759 · Python · MIT · updated Apr 2026

Supercharge BigQuery with BigFunctions

BigFunctions is a library of 150+ BigQuery remote functions (Python, JS, SQL) plus a CLI tool (`bigfun`) for deploying and managing them in your own GCP project. It covers a genuinely wide range: text transforms, geo operations, ML utilities, AI calls, notifications, and data loading from SaaS sources. Data engineers who live in BigQuery and wish it could do more without leaving SQL will get the most out of it.

The public deployment model is clever — you can call `bigfunctions.eu.faker(...)` directly without installing anything, which lowers the barrier to trying it considerably. The YAML-driven function definition format makes contributing new functions straightforward and keeps the catalog consistent. Coverage of practical gaps in BigQuery is genuinely useful: H3 indexing, Prophet forecasting, IP geolocation, and GA4 unnesting are all things people regularly build themselves. The `run_python` escape hatch is useful when you need arbitrary Python inside a SQL pipeline.

The public functions depend on Unytics's infrastructure staying up and the Cloud Run endpoints remaining available — there's no SLA, and if they go down your queries break silently. Calling external HTTP endpoints from inside BigQuery queries means latency and cost surprises at scale; a function like `geocode` or `ask_ai` per row is going to hurt on large tables. The `wip/` folder contains things like `generate_website.yaml` that look abandoned and make it hard to tell what's actually production-ready versus an experiment. There's no versioning strategy for the public functions, so a breaking change in a function definition affects every caller immediately.

View on GitHub → Homepage ↗

// want more like this?

We dig through GitHub every week and send a few repos picked for what you actually care about — each with an honest take like this one.

Get finds in your inbox → Search again →