Streaming data · June 2026 Academic lab

Real-time auctions

Real-time ingestion of auction events: external API → Cloud Run → Pub/Sub and BigQuery → Looker Studio dashboards.

Report · on request

$ 12.45012.90013.35014.10012.450 ▲

API → Cloud Run → Pub/Sub → BigQuery

What it is

An event-driven pipeline: a Cloud Run function receives each bid over HTTP, normalizes it, publishes it to Pub/Sub and lands it in BigQuery to query with SQL and explore in Looker Studio.

5Google Cloud services

What I built

A Python HTTP endpoint that validates and normalizes the JSON
A Pub/Sub topic and subscription
An analytical table in BigQuery and business queries
Interactive dashboards with filters and periodic refresh

The hard parts

Plan B, live

Dataflow didn’t come up in the lab environment: loading was handled from Cloud Run with insert_rows_json, without breaking real time.