I'm unhappy with using Postgres as a data warehouse

That’s a pretty common pain point. Postgres is great as an operational database, but it often starts to hurt as a warehouse when data volume, query concurrency, or analytics complexity grows.

A few reasons people move away from Postgres for warehousing:

• Columnar performance: Postgres is row-oriented, so large scans and aggregations are usually slower than columnar engines.
• Concurrency: lots of BI users or dashboard refreshes can strain it.
• Transform workload: heavy ETL/ELT can compete with production usage if it’s sharing the same system.
• Storage/partition management: large fact tables and retention policies can get cumbersome.
• Cost scaling: you often end up scaling up a single big Postgres instance rather than scaling out analytically.

Common alternatives depend on what you need:

• BigQuery / Snowflake / Redshift: managed cloud warehouses, easiest path if you want strong analytics performance without operating infrastructure.
• ClickHouse: very fast for analytical queries, great if you want speed and are okay with a more hands-on system.
• DuckDB: excellent for local/embedded analytics or small-to-medium pipelines.
• Databricks / Spark-based lakehouse: useful if your warehouse also needs large-scale data engineering and ML.
• Open-source warehouse/lakehouse stack: e.g. S3 + Iceberg/Delta + Trino, if you want more control and lower vendor lock-in.

If you want, I can help you choose a replacement based on:

1. data size,
2. number of users / queries,
3. cloud provider,
4. budget,
5. whether you need SQL BI, streaming, or ML.

If you answer those, I can recommend a better fit than Postgres.