Add sample 'orders-streaming-analytics'

README.md

@@ -19,7 +19,7 @@ See respective README files in sub-directories for details.
 - [LoR](lor/README.md): Using Dataflow, Cloud Run and Spanner to explore Lord of the Rings characters.
 - [Network Digital Twin](telco-and-csp/README.md): Advanced usage of graph to Spanner Graph to model, visualize, and query a complex telecommunications network.
 - [Transit Fraud Detector](TransitFraud/README.md): Advanced usage of graph capabilities to detect fraud.
-
+- [Orders Streaming Analytics](orders-streaming-analytics/README.md): Dual-path streaming with Spanner: Precomputed Lakehouse aggregates and Ad-Hoc analytics via Data Boost.
 
 ## Notebooks
 Some of these notebooks are hosted in external Google Cloud repositories.

orders-streaming-analytics/.gitignore

@@ -0,0 +1,4 @@
+src/java-tpch-stream-generator/target
+output
+checkpoints
+.env

orders-streaming-analytics/README.md

@@ -0,0 +1,27 @@
+# Spanner DataBoost Streaming Demo
+
+This demo implements a dual-path streaming architecture ingesting orders (TCP-H dataset) where each branch addresses a distinct analytical goal.
+
+The left branch (Lakehouse Aggregates path) handles predictable, day-to-day reporting by precomputing cumulative aggregates, such as daily order totals, and saving them directly into Spanner aggregate tables only when relevant data changes.
+
+The right branch (Direct Spanner path) continuously maintains the raw orders table to support ad-hoc queries not covered by precomputations. By leveraging Spanner Data Boost alongside its columnar engine, these ad-hoc queries run efficiently without impacting the primary Spanner instance, thereby eliminating the need to over-provision it.
+
+![DataBoost Streaming Demo](SpannerStreamingLakehouse-Dataflow.drawio.png)
+
+### System requirements
+
+- gcloud cli
+- uv https://docs.astral.sh/uv/#installation (for dependencies management)
+- jq https://jqlang.org/ (for parsing gcloud json response reliably, used by the bash scripts during the notebook)
+
+
+## Setup
+
+The entire demo architecture notes, GCP infrastructure provisioning, Spark job sources, producer execution, and dashboard/ad-hoc query code is in [`demo.ipynb`](demo.ipynb).
+
+```bash
+uv sync
+uv run jupyter lab
+```
+
+Open `demo.ipynb` and run the cells top to bottom.