Trust Center

We built Lighthouse to operate safely inside your data stack - with minimal access, full transparency, and data residency you control. Your warehouse stays your warehouse. We only retrieve what’s needed to evaluate your metrics — nothing more.

How Lighthouse Handles Your Data

Lighthouse connects to your data warehouse using a read-only, scope-limited role.
Two types of access occur:
1) Schema Metadata Discovery (no row-level data)
Used to assist in metric setup and autocomplete:
- Table & view names
- Column names & data types
- Schema structure
* No row-level or cell-level records are accessed during this step.
2) Metric Execution (result sets only)
- Lighthouse executes only the SQL you define
- Only the result set from that query is retrieved
- If metric is aggregated → only aggregates returned
- If metric is raw-level → raw-level subset returned
* We never replicate or crawl underlying raw datasets.

Architecture at a Glance

Your Data Warehouse (Snowflake/BigQuery/etc)
                       ↓ read-only, scope-limited role
Lighthouse executes your SQL
                        ↓
Only result sets + metadata returned
                        ↓
Encrypted storage for history & UI display

Least-Privilege Access Model

Permission

Purpose

USAGE on schema

Metadata discovery

SELECT on specific tables/views

Metric execution

information_schema access

Column metadata lookup

No WRITE/DML privileges

Enforced by design

* Access can be revoked at any time by removing the role.

Data Residency

All data processed by Lighthouse is stored in the United States using U.S.-based infrastructure providers.

Data Retention & Lifecycle

Stage

Policy

During service

Metric results & metadata stored securely for history

Offboarding

All retained data deleted within 30 days

Access revocation

Customer can revoke warehouse access instantly

Access Controls

Lighthouse enforces:
- SSO via Google Workspace
- MFA for privileged access
- Principle of least privilege
- Quarterly access reviews
- 24-hour offboarding window

Encryption

Category

Standard

In transit

TLS 1.2+

At rest

AES-256

Logging & Monitoring

We monitor:
- Authentication events
- Infrastructure & deployment activity
- Metric execution telemetry Alerts feed into our incident response procedures.

Reliability & Continuity

Objective

Target

RTO

24 hours

RPO

24 hours

Subprocessors

Lighthouse uses a minimal set of U.S.-based infrastructure providers that meet SOC 2 or equivalent security standards.
A full list is available to customers upon request.
📩 Request list: admin@lighthouseai.co

High-Level Preview:

Category

Purpose

Application Hosting

Secure cloud infrastructure

Authentication

Identity & SSO

Data Storage

Encrypted persistence

Monitoring

Uptime & performance

Collaboration / Controls

Internal security governance

Compliance

Lighthouse AI is aligned with the SOC 2 Trust Services Criteria for Security, Availability, and Confidentiality. We are formally progressing toward SOC 2 Type II certification.

Security FAQ

Q: Does Lighthouse have access to my full data warehouse?
No. Access is read-only and scoped only to tables/views used in metrics.
Q: Do you store raw data?
No. Only metric result sets and evaluation metadata are stored.
Q: What happens if we disconnect?
All stored data is deleted within 30 days and access is immediately revoked.
Q: Do you support least-privilege?
Yes — role scope is fully controlled by the customer.
Q: Are subprocessors U.S.-based?
Yes — and a full list is available upon request.

Have additional security questions? 📩 admin@lighthouseai.co