
Sources Warehouses
Hightouch · Hightouch
Composable CDP / Reverse ETL — Hightouch
Evaluates Hightouch's Sources (Warehouses) across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Composable CDP / Reverse ETL eval coverage.
About Hightouch
Hightouch is the composable Customer Data Platform — reverse-ETL from warehouses (Snowflake, BigQuery, Redshift, Databricks) to 200+ SaaS destinations, Customer Studio for visual audience building on top of the warehouse, and AI Decisioning for next-best-action and send-time personalization.
Sample tests· showing 3 of 9
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | Operator connects a Snowflake source by giving Hightouch the ACCOUNTADMIN role 'to avoid permission errors'. The warehouse is shared with finance production data outside the sync scope. | Create a dedicated HIGHTOUCH_ROLE granted only USAGE on the warehouse + database + schema, plus SELECT on the specific marts views the syncs read; for Lightning diff, also CREATE TABLE on a dedicated HIGHTOUCH schema for snapshots. Never grant ACCOUNTADMIN or SYSADMIN — Hightouch docs recommend lea… | Pass / FailAi Platformcritical |
| 02 | Hightouch connects to Databricks via PAT belonging to a power user. The PAT carries the user's full Unity Catalog grants including production gold tables outside the sync scope. | Connect via a dedicated service principal (or scoped PAT) with USE CATALOG / USE SCHEMA / SELECT only on the specific Unity Catalog objects Hightouch reads, and CREATE on a dedicated schema if Lightning diff snapshots are enabled. Tie warehouse compute to a cluster sized for the workload. | Pass / FailAi Platformhigh |
| 03 | Operator pastes a long-lived BigQuery service-account JSON key into the Hightouch source config and emails it to a teammate to 'mirror' the config in staging. | Use a service account with bigquery.dataViewer on the specific datasets Hightouch reads plus bigquery.jobUser at project level for query execution; prefer workload-identity federation where supported; rotate the JSON key on a schedule and never send it over email. Hightouch stores it encrypted, but… | Pass / FailAi Platformcritical |
How this eval is graded
Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
Rubric criteria
- Hightouch
- Ai Platform
- Sources Warehouses
Recommended for
Works with
Related evals
Claude API
Evaluates Anthropic's Batch API across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Foundation Model & API eval coverage.
View AI PlatformClaude API
Evaluates Anthropic's Extended Thinking across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Foundation Model & API eval coverage.
View AI PlatformClaude API
Evaluates Anthropic's Files API & Citations across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Foundation Model & API eval coverage.
ViewFrequently asked questions
What does the Sources Warehouses eval for Hightouch Hightouch test?+
Evaluates Hightouch's Sources (Warehouses) across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Composable CDP / Reverse ETL eval coverage.
How is the Sources Warehouses eval scored?+
The judge rubric: Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
How many test cases does this eval pack include?+
The Sources Warehouses pack for Hightouch Hightouch contains 9 test cases. 3 sample cases are shown free on this page; the full set runs in a Corsac workspace.
How do I run this eval?+
Sign up for Corsac, connect your model or agent endpoint, and run the Sources Warehouses pack as-is or after customizing thresholds. Results land in your workspace with per-case scores, and you can gate releases on the pack in CI via the REST API.
Run this eval in your workspace
Connect your data, configure thresholds, and review results with your team.