What Is Amazon Redshift?
Amazon Redshift is AWS’s cloud data warehouse service, designed for large-scale data analytics. Launched in 2012, it was one of the first cloud data warehouses and remains popular for organizations heavily invested in the AWS ecosystem.
Key characteristics:
- Columnar storage for fast analytics
- Massively parallel processing (MPP)
- Tight AWS integration
- Multiple cluster configurations
- Serverless option available
- Mature, proven platform
Redshift Architecture
Cluster-Based
Traditional Redshift uses clusters:
- Leader node (query coordination)
- Compute nodes (data storage and processing)
- Node types: Dense Compute (DC) or RA3
Redshift Serverless
Newer serverless option:
- No cluster management
- Automatic scaling
- Pay for compute used
- Simpler pricing
Columnar Storage
- Data stored by column, not row
- Efficient compression
- Fast analytical queries
- Zone maps for query optimization
Redshift for Finance
Common financial use cases:
Financial consolidation
- Multi-entity data centralization
- Historical financial data
- Trend analysis
Operational reporting
- Real-time dashboards
- KPI tracking
- Operational metrics
Customer analytics
- Revenue analysis
- Customer segmentation
- Lifetime value
Cost analysis
- Spend analytics
- Vendor analysis
- Cost allocation
Redshift vs. Other Data Warehouses
| Aspect | Redshift | Snowflake | BigQuery |
|---|---|---|---|
| Cloud | AWS only | Multi-cloud | GCP only |
| Architecture | Cluster/Serverless | Virtual warehouses | Serverless |
| Pricing | Per-node or serverless | Per-second | Per-query |
| AWS integration | Native | Via PrivateLink | N/A |
| Scaling | Resize cluster | Instant | Automatic |
| Concurrency | Scaling feature | Per-warehouse | High |
Redshift Features
Redshift Spectrum
- Query data in S3 directly
- No loading required
- Extend warehouse to data lake
Concurrency Scaling
- Handle query spikes
- Automatic additional capacity
- Pay only when used
Materialized Views
- Pre-computed query results
- Automatic refresh
- Query acceleration
Federated Query
- Query across Redshift, RDS, S3
- Single SQL statement
- Real-time data access
ML Integration
- Redshift ML
- SageMaker integration
- SQL-based predictions
Redshift Pricing
Provisioned clusters:
- Per-node per-hour
- Reserved instances for discounts
- Spectrum charges for S3 queries
Serverless:
- Redshift Processing Units (RPU)
- Per-second billing
- No idle costs
Storage:
- Managed storage pricing
- Automatic tiering
How Go Fig Works with Redshift
Go Fig integrates with Redshift for financial analytics:
Redshift as source:
- Connect to Redshift clusters
- Query tables and views
- Use existing data models
Redshift as destination:
- Load data from ERPs, CRMs
- Build unified data layer
- Support broader analytics
Go Fig value-add:
- Semantic layer over Redshift
- Excel delivery without SQL
- Combine with non-AWS sources
- AI-powered insights
Redshift Best Practices
Performance
- Choose appropriate distribution keys
- Define sort keys for common queries
- Use compression encodings
- Vacuum and analyze regularly
Cost Management
- Right-size clusters
- Use reserved instances for steady workloads
- Leverage Spectrum for cold data
- Consider serverless for variable workloads
Security
- VPC isolation
- Encryption at rest and in transit
- IAM integration
- Column-level access control
Getting Started with Redshift
For finance teams considering Redshift:
- Assess AWS investment: Already on AWS? Natural fit.
- Choose deployment: Provisioned vs. Serverless
- Plan data loading: S3, DMS, Fivetran, etc.
- Design schema: Distribution, sort keys, compression
- Select access tools: QuickSight, Go Fig, third-party BI
Go Fig provides a business-friendly layer, making Redshift data accessible through Excel and dashboards without requiring SQL skills or deep AWS knowledge.