Hands-on ClickHouse labs, grouped into series — from a 90-minute analytics sprint to a full day migrating a warehouse workload. Each runs on your own free ClickHouse Cloud trial.
Series 012 workshops
From empty service to live app.
Full end-to-end builds. Start with nothing but a ClickHouse Cloud service and your own AI coding agent, and finish with a real-time application you can demo. Longest format, highest ceiling.
In one three-hour sitting, take a real analytics app live on ClickHouse Cloud with your own AI coding agent — real-time CDC, conversational BI, observability, an AI SRE, and traced AI chat.
Open workshop →Stream public prediction-market data into ClickHouse Cloud, maintain one-minute aggregates, investigate movement, put an AI agent on the live tables, and publish a market pulse dashboard.
Open workshop →Series 025 workshops
Same workload, new engine.
Move an existing warehouse workload to ClickHouse decision by decision — schema, ingestion, pipeline, benchmark — on infrastructure you stand up and tear down yourself.
A 6-hour hands-on lab: stand up a production-shaped Snowflake workload, plan the migration decision by decision, move 50M rows to ClickHouse Cloud, rebuild the dbt pipeline on MergeTree engines, and benchmark all seven queries end to end.
Open workshop →A 4-hour hands-on lab: use MigrationRoom's AI-assisted workflow to move a Databricks TPC-H workload, prove Unity Catalog and a second REST catalog can stay zero-copy, and build a measured ClickHouse-native hot path.
Open workshop →Migrate a production-shaped observability stack from Elasticsearch, Kibana, Filebeat, and Elastic APM to ClickHouse Cloud, OpenTelemetry, and HyperDX. Design the schema, validate live parity, and complete a controlled cutover.
Open workshop →Move a live e-commerce workload off AWS RDS Postgres onto ClickHouse Managed Postgres with a minute of write downtime, stream the analytical copy into ClickHouse with ClickPipes, and route the dashboard's queries to it without changing a line of SQL.
Open workshop →Three hours: export a BigQuery public dataset, query it in place from object storage, then take the schema BigQuery hands you and beat it three times over -- on compressed bytes, on rows read for a lookup, and on a per-minute conversion funnel no sort key can rescue.
Open workshop →Series 033 workshops
One capability, one sitting.
Focused labs that go deep on a single ClickHouse capability instead of a whole application. Self-contained and short enough for a lunch slot — a good first workshop, or a follow-up to a BUILD.
A 90-minute hands-on lab: load ~26.5M forex ticks from object storage, run real market-analytics queries, and try the built-in AI Assistant and Agents. Optional live dashboard on your own data.
Open workshop →A 90-minute hands-on lab: put six models and three prompt strategies into a contest for natural-language-to-SQL over ClickHouse, grade them on execution accuracy, and pick what to ship by cost per correct answer — with Langfuse in the loop from step one.
Open workshop →This is a step-by-step Langfuse workshop built on a small TypeScript sample application: the Dad IT Support Agent. The workshop covers the full AI engineering loop with Langfuse: tracing, prompt management, monitoring, datasets, experiments, and evaluation.
Open workshop →