Skip to content

Lightdash Cloud Setup

On this page, you will:

  • Create a Lightdash Cloud account and project
  • Connect Lightdash to your GitHub dbt-transform repository
  • Connect Lightdash to Snowflake using the SVC_LIGHTDASH service account
  • Understand how Lightdash reads your dbt project and discovers models
  • Configure deployment and development environments
  • Verify Lightdash can query your dbt models

Overview

Lightdash Cloud is the managed SaaS version of Lightdash. It provides: - Zero infrastructure management (no ECS, RDS, or Docker to configure) - Automatic updates and security patches - Hosted at https://your-org.lightdash.cloud - $2400/month flat rate (unlimited users)

This page covers the quick-start path using Lightdash Cloud. If you prefer self-hosting, skip to Self-Hosted Lightdash.

┌─────────────────────────────────────────────────────────────────────────┐
│                     LIGHTDASH CLOUD ARCHITECTURE                        │
├─────────────────────────────────────────────────────────────────────────┤
│                                                                         │
│  GitHub                  Lightdash Cloud              Snowflake         │
│  ──────                  ────────────────             ─────────         │
│                                                                         │
│  ┌──────────────┐        ┌──────────────┐            ┌──────────────┐  │
│  │ dbt-transform│───────▶│ Lightdash    │───────────▶│ ANALYTICS    │  │
│  │ repository   │        │ • Reads dbt  │            │ • REPORTING  │  │
│  │              │        │ • Compiles   │            │   schema     │  │
│  │ • models/    │        │ • Syncs      │            └──────────────┘  │
│  │ • metrics    │        │   metrics    │                             │
│  └──────────────┘        └──────────────┘            SVC_LIGHTDASH     │
│                                                       (read-only)       │
│  Lightdash clones your dbt repo, reads YAML,                           │
│  and queries Snowflake to build dashboards.                            │
│                                                                         │
└─────────────────────────────────────────────────────────────────────────┘

Sign Up for Lightdash Cloud

  1. Navigate to Lightdash Cloud
  2. Click Start for free or Get started
  3. Sign up with:
  4. Email (or Google/GitHub SSO)
  5. Organisation name (your company name, e.g., "Acme Data")

  6. Project name (e.g., "Analytics")

  7. Complete the onboarding wizard

You'll land on the Lightdash home page with a prompt to connect your dbt project.

Pricing Confirmation

Lightdash Cloud is $2400/month (unlimited users). A 14-day free trial is available. No credit card required to start the trial.

Connect Your GitHub Repository

Lightdash reads your dbt project directly from GitHub. It clones the repository, parses dbt_project.yml and model YAML files, and discovers metrics and models.

Step 1: Authorise GitHub Access

  1. In Lightdash Cloud, click Connect project
  2. Select GitHub as your repository provider
  3. Click Authorise Lightdash to install the Lightdash GitHub App
  4. Select your organisation (the one containing the dbt-transform repo)
  5. Choose Only select repositories and select dbt-transform
  6. Click Install & Authorise

Step 2: Configure the Project

After authorising GitHub, Lightdash prompts you to configure the project:

Field Value
Repository your-org/dbt-transform
Branch main (production branch)
Project subdirectory Leave blank (unless dbt project is in a subfolder)
Target prod (from profiles.yml)

Lightdash will clone the repository and attempt to parse the dbt project.

Multiple Environments

You can create multiple Lightdash projects for different branches (e.g., main for production, develop for staging). This mirrors dbt Cloud's environment model.

Connect to Snowflake

Lightdash needs Snowflake credentials to query your data. Use the SVC_LIGHTDASH service account created in the previous step.

Step 1: Retrieve Credentials from AWS Secrets Manager

Fetch the credentials you stored earlier:

aws secretsmanager get-secret-value \
    --secret-id "lightdash/snowflake-credentials" \
    --query SecretString \
    --output text \
    --profile infrastructure-admin | jq -r .

Copy the values for: - account - user - role - warehouse - database - schema - private_key

Step 2: Configure Snowflake Connection in Lightdash

In Lightdash Cloud, navigate to Settings → Project settings → Data connection:

Field Value
Warehouse type Snowflake
Account your-account.snowflakecomputing.com (from secrets)
User SVC_LIGHTDASH
Role SVC_LIGHTDASH
Database ANALYTICS
Warehouse REPORTING
Schema REPORTING

For authentication, choose Key pair:

Field Value
Authentication method Key pair
Private key Paste the entire private key from AWS Secrets Manager (including -----BEGIN RSA PRIVATE KEY----- and -----END RSA PRIVATE KEY-----)

Click Test connection.

Expected result: 

✅ Connection successful!
Connected to ANALYTICS.REPORTING as SVC_LIGHTDASH.

Click Save.

Private Key Security

The private key is stored encrypted in Lightdash Cloud. Lightdash uses it to authenticate to Snowflake but never exposes it to users.

Troubleshooting Connection Issues

Error: "Authentication failed" - Verify the private key is correct (copy entire key including headers) - Ensure public key was added to Snowflake via Terraform - Check the user exists: SHOW USERS LIKE 'SVC_LIGHTDASH';

Error: "Insufficient privileges" - Verify SVC_LIGHTDASH has ANALYTICS_REPORTER granted: SHOW GRANTS TO ROLE SVC_LIGHTDASH; - Test access manually: USE ROLE SVC_LIGHTDASH; SELECT COUNT(*) FROM analytics.reporting.fct_exchange_rates;

Error: "Database not found" - Ensure database is ANALYTICS (not ANALYTICS_DEV) - Verify schema is REPORTING (not MARTS or STAGING)

Compile and Sync the dbt Project

After connecting Snowflake, Lightdash compiles your dbt project to discover models and metrics.

Step 1: Trigger Initial Compile

In Lightdash, navigate to Settings → Project settings → dbt:

  1. Click Compile project
  2. Lightdash will:
  3. Clone the dbt-transform repository
  4. Run dbt deps to install packages
  5. Run dbt compile to generate the manifest.json
  6. Parse model YAML files to discover metrics

This takes 30-90 seconds.

Expected result: 

✅ Compilation successful!
Discovered 5 models, 12 metrics, 8 dimensions.

Step 2: Review Discovered Models

Navigate to Explore in the left sidebar. You should see:

  • fct_exchange_rates (from models/marts/core/fct_exchange_rates.sql)
  • dim_products (from models/marts/core/dim_products.sql)
  • Other models you've published to the REPORTING schema

Click a model to see: - Dimensions (columns from the model) - Metrics (defined in the model's YAML file, e.g., fct_exchange_rates.yml) - Joins (if you've defined relationships in dbt)

How Lightdash Discovers Models

Lightdash reads your dbt project and looks for: 1. Models in models/ directory 2. YAML files with meta tags for metrics and dimensions 3. Relationships defined in sources.yml and model YAML

It does not create any tables or views in Snowflake — it only reads existing dbt models.

Understanding Lightdash Environments

Lightdash has two environment concepts:

Production Environment (Main Branch)

  • Reads dbt models from the main branch
  • Queries the production Snowflake database (ANALYTICS.REPORTING)
  • Used by end users for dashboards and reports
  • Syncs automatically when you merge PRs to main

Development Environment (Feature Branches)

  • Developers can create "developer previews" from feature branches
  • Queries the development Snowflake database (ANALYTICS_DEV.ANALYTICS_DEV_REPORTING)
  • Used for testing new metrics before merging to main
  • Isolated from production

To create a development environment:

  1. Navigate to Settings → Environments
  2. Click + Add environment
  3. Select branch (e.g., feature/new-metrics)
  4. Configure Snowflake to use ANALYTICS_DEV database
  5. Compile the project

Developer Workflow

  • Create a branch in dbt-transform with new metrics
  • Create a Lightdash dev environment for that branch
  • Test metrics in Lightdash before merging
  • Merge to main → Lightdash production auto-syncs

Auto-Sync on Git Push

Lightdash can automatically recompile when you push to GitHub:

  1. Navigate to Settings → Project settings → Git integration
  2. Enable Auto-sync on push
  3. Select branches to sync: main (and optionally develop)

Now when you: - Merge a PR to main → Lightdash automatically recompiles and syncs new metrics - Push to develop → Lightdash updates the development environment

This means your BI tool stays in sync with your dbt project automatically.

Test Querying a Model

Step 1: Navigate to Explore

  1. Click Explore in the left sidebar
  2. Select fct_exchange_rates (or another model)

Step 2: Build a Query

Lightdash provides a drag-and-drop interface to build queries:

  1. Dimensions (left sidebar): Select rate_date, target_currency
  2. Metrics (left sidebar): Select a metric (if defined in YAML, e.g., average_exchange_rate)
  3. Filters: Add filter base_currency = GBP
  4. Click Run query

Lightdash generates SQL and queries Snowflake:

-- Generated by Lightdash
SELECT
    rate_date,
    target_currency,
    AVG(exchange_rate) AS average_exchange_rate
FROM analytics.reporting.fct_exchange_rates
WHERE base_currency = 'GBP'
GROUP BY rate_date, target_currency
ORDER BY rate_date DESC;

Results appear in a table. Click Chart to visualise.

Step 3: Create a Chart

  1. Click Chart (top-right)
  2. Select chart type: Line chart
  3. Configure:
  4. X-axis: rate_date
  5. Y-axis: average_exchange_rate
  6. Group by: target_currency

You now have a line chart showing exchange rate trends by currency.

Step 4: Save the Chart

  1. Click Save chart
  2. Name it: "GBP Exchange Rates Trend"
  3. Choose a space (or create one): "Exchange Rates"
  4. Click Save

The chart is now saved and can be added to dashboards or shared with teammates.

First Chart Created

You've successfully queried a dbt model and created a chart in Lightdash. This chart is dynamic — as your dbt models update, the chart updates automatically.

Verify the Setup

Checklist to confirm Lightdash is working:

  • GitHub repository connected (Lightdash can clone dbt-transform)
  • Snowflake connection successful (Lightdash can query ANALYTICS.REPORTING)
  • dbt project compiled (Lightdash discovered your models and metrics)
  • At least one model visible in Explore (e.g., fct_exchange_rates)
  • Test query runs successfully (Lightdash can fetch data from Snowflake)
  • Chart created and saved (Lightdash can visualise data)

If all checks pass, Lightdash is ready to build dashboards.

Cost Recap

Lightdash Cloud: - $2400/month (unlimited users) - 14-day free trial - No infrastructure costs (managed by Lightdash)

Snowflake compute: - ~$25/month (estimated, based on 100 dashboard views/day) - Uses REPORTING warehouse (X-Small, auto-suspend 60 seconds)

Total monthly cost: ~$2425 (Lightdash Cloud + Snowflake compute)

For cost-conscious teams, Self-Hosted Lightdash reduces this to ~$55/month (infrastructure + compute).

Alternative: Self-Hosted Lightdash

If $2400/month is beyond your budget, Lightdash offers a self-hosted option:

  • Deploy Lightdash to AWS ECS (infrastructure cost: ~$30-50/month)
  • Same features as Cloud (minus managed hosting)
  • You manage updates, backups, and scaling
  • Terraform deployment (infrastructure as code)

See Self-Hosted Lightdash for setup instructions.

Summary

You've set up Lightdash Cloud:

  • Created a Lightdash Cloud account and project
  • Connected to GitHub (dbt-transform repository)
  • Connected to Snowflake (SVC_LIGHTDASH service account, REPORTING schema)
  • Compiled the dbt project and discovered models
  • Configured auto-sync on Git push (Lightdash stays in sync with dbt)
  • Created and saved a test chart
  • Understood production and development environments

Lightdash is now reading your dbt project and querying Snowflake. The next step is to connect your dbt project properly and define metrics in your dbt YAML files.

What's Next

If you want to self-host instead of using Lightdash Cloud, deploy Lightdash to AWS ECS for cost savings.

Otherwise, skip to Connect dbt Project to configure how Lightdash discovers your dbt models and metrics.

Continue to Self-Hosted Lightdash → (optional)

Continue to Connect dbt Project → (if using Cloud)