Amazon Redshift is the data warehouse of choice for organisations running on AWS. It handles petabytes of data, integrates natively with the broader AWS ecosystem (S3, Glue, SageMaker, QuickSight), and delivers fast query performance for complex analytical workloads. Loading invoice and AR data into Redshift makes it queryable alongside all your other business data.
Connection setup
TallyArc connects to Redshift using the standard Redshift connector over JDBC/ODBC. You'll need:
- Your cluster endpoint (e.g.
cluster.abc.us-east-1.redshift.amazonaws.com) - Port (default 5439)
- A dedicated database user with
CREATE TABLEandINSERTon your target schema - Username and password for that user
Ensure your Redshift cluster's security group allows inbound connections from TallyArc's egress IP range (available in your TallyArc account settings under Data Platforms).
Recommended schema design
For Redshift, organise invoice data in a star schema:
- fact_invoices — one row per invoice with all numeric measures (amounts, days to pay, line item count)
- fact_payments — one row per payment event
- dim_clients — client attributes (industry, size, location, payment terms)
- dim_date — a date dimension table for time-based aggregations
This structure queries efficiently in Redshift's columnar storage format and works out of the box with BI tools like Amazon QuickSight, Tableau, and Looker.
Connecting in TallyArc
- Go to Data → Amazon Redshift → Connect
- Enter cluster endpoint, port, database name, username, and password
- TallyArc will create the required tables on first sync
- Run an initial full load, then configure incremental syncs going forward
AWS integrations
With invoice data in Redshift, AWS-native analytics become straightforward:
- Amazon QuickSight — build AR dashboards with native Redshift connectivity; SPICE caching gives sub-second query response
- AWS Glue — schedule ETL jobs to transform raw invoice data into reporting-ready aggregations
- Amazon SageMaker — build payment prediction models using Redshift ML (SQL-based model training in-database)