Project
LandInvest on Google Cloud
The data platform and Market Explorer application behind LAISER's land-investment research, built as a data lake, BigQuery layer, serving export, and Cloud Run web app with pipeline-only deploys.
- Status
- active
- Started
- Source
- Private repository
LandInvest is the engineering counterpart to the Land Alpha research programme: the place where data pipelines, warehouses, and a user-facing application are built to production standards. It began in 2026 as a research platform (a county-level boom-onset classifier with scoring and dashboards) and in October 2026 was re-based as a Google Cloud code base.
What it is
- A data lake and warehouse. Ingestion and curation jobs land raw sources, and a BigQuery layer exposes curated and serving datasets.
- A serving layer. An exported, versioned serving bundle that the application can load without touching production credentials.
- Market Explorer. A FastAPI web application for exploring county-level markets, served from Cloud Run.
How it ships
Two protected branches: dev for integration and main for what is deployed. Every merge
request runs lint, generated-SQL freshness checks, and unit tests on a project-owned runner in
GCP. A release is a fast-forward merge to main, a semantic version tag, and a GitLab Release
whose notes are the matching changelog section. Deploys happen only through the pipeline; nothing
a developer runs locally can reach Cloud Run.