data-tier
Lakehouse, native to basecamp.
data-tier is the /root take on a lakehouse — but built entirely on OSS primitives: MinIO for object store, Apache Iceberg for table format, Trino for SQL, and Flink for streaming. Not Databricks. Every component is swappable.
What you can do
Iceberg tables, OSS
ACID, time-travel, hidden partitioning — all on top of MinIO or S3.
Trino SQL layer
Query Iceberg, Postgres, and Kafka from the same SELECT statement.
Streaming ingestion
Flink pipelines write into the same Iceberg tables batch reads from.
Schema evolution
Add columns, rename them, backfill defaults — the reader survives.
One catalog
Hive Metastore–compatible, so any lakehouse client works with data-tier.
Not Databricks
Every component is OSS. Every substrate choice is a swap away.
Explore the rest of /root
The 9-tier K8s-native platform
Reproducible experiments
Observability & tracing
On-call alerting
Cloud infrastructure as code
Platform control CLI
LLM routing & rate limits
MCP tool server plans
AI-augmented operations
Weekly logs + runbooks
Get started with data-tier
Clone the repo, read the plan, and start building your own version.
All projects