An accelerated lakehouse for every customer, team, and agent
# spicepod.yamldatasets: - from: s3://spiceai-demo-datasets/taxi_trips/2024/ name: taxi_trips params: { file_format: parquet, s3_auth: public } - from: postgres:taxi_zones name: taxi_zones params: pg_host: localhost pg_port: 15432 pg_db: taxi pg_user: postgres pg_pass: ${secrets:SPICE_PG_PASS} pg_sslmode: disablesql> SELECT z.zone, COUNT(*) AS trips FROM taxi_trips t -- S3 JOIN taxi_zones z -- PostgreSQL ON t."PULocationID" = z.location_id GROUP BY z.zone ORDER BY trips DESC LIMIT 3;JFK Airport 145240Midtown Center 143471Upper East Side South 142708Time: 2.705 seconds# spicepod.yaml: add to each datasetacceleration: enabled: true$ spice runLoaded 265 rows (30.46 kiB) for dataset taxi_zonesLoaded 2,964,624 rows (399.38 MiB) for dataset taxi_tripssql> -- same query, served from the local copy SELECT z.zone, COUNT(*) AS trips FROM taxi_trips t JOIN taxi_zones z ON t."PULocationID" = z.location_id GROUP BY z.zone ORDER BY trips DESC LIMIT 3;JFK Airport 145240Midtown Center 143471Upper East Side South 142708Time: 0.026 seconds# spicepod.yaml: index the zone column, add a modeldatasets: - from: postgres:taxi_zones name: taxi_zones acceleration: enabled: true columns: - name: zone full_text_search: enabled: true embeddings: - from: embeddings embeddings: - name: embeddings from: openai:text-embedding-3-small params: openai_api_key: ${secrets:OPENAI_API_KEY}models: - name: claude from: anthropic:claude-sonnet-4-5 params: anthropic_api_key: ${secrets:SPICE_ANTHROPIC_API_KEY}sql> SELECT zone, borough FROM rrf(text_search(taxi_zones, 'airport'), vector_search(taxi_zones, 'where flights land')) ORDER BY _fused_score DESC LIMIT 3;# example outputJFK Airport QueensLaGuardia Airport QueensNewark Airport EWRsql> SELECT zone, ai(concat_ws(' ', 'Categorize the zone', zone, 'in a single word.')) AS category FROM taxi_zones ORDER BY location_id LIMIT 3;# example outputNewark Airport AirportJamaica Bay WildlifeAllerton/Pelham Gardens ResidentialOne spicepod.yaml file defines the lakehouse: SQL federation across your data sources, plus acceleration, search, and inference. The Spice Cloud API deploys it for each customer, internal team, or agent.
Every deployment runs the same standard lakehouse, so operating hundreds stays simple. Each can still use its own data sources and settings. A SaaS company can give each customer its own lakehouse; an enterprise data platform team can do the same for each internal team.
A coding agent can write and run the spicepod for you in minutes.
