Case study · Retail pharmacy · Analytics
Helped India's 3rd largest pharmacy chain boost sales effectiveness with a Redshift data warehouse.
Retail pharmacy and wellness · Omnichannel
Data across store databases, sheets, CRM and transaction systems, with 3x year-on-year growth and no unified place to analyse it.
30 min
Data processing, down from 6 hours
80%
Improvement in decision-making
300+
Tables across 5+ data sources
The opportunity
Five systems holding the answer, and no way to ask them together.
This omnichannel pharmacy chain carries over 100,000 products across a hyperlocal 24x7 network. Its data lived across store databases, spreadsheets, CRM and transaction systems-with no unified platform to drive analytics from any of it.
Combining those sources to make a decision was difficult, which meant sales performance, marketing campaigns and inventory were all managed with partial visibility. A 3x year-on-year increase in data volume made that harder every year.
No unified analytics platform
Data sat across store databases, sheets, CRM and transaction systems.
Sources could not be combined
Data-driven decisions on sales, campaigns and inventory were difficult to make.
3x data growth year on year
Rising volume made decision-making more complex, not less.
No campaign-to-sales visibility
Performance by brand, store and manufacturer could not be seen together.
The impact
One warehouse, fifteen dashboards, thirty-minute processing.
Centralised on Redshift
A single warehouse where data from every source resides for analysis.
Self-serve ETL
A pipeline on Managed Airflow, EMR Serverless, S3 and Redshift covers 300+ tables from 5+ sources.
15+ dashboards on Superset
Sales by category and store, plus discount and purchase analysis, in one place.
Six hours to thirty minutes
Complex analytical queries run against consolidated data rather than across systems.
The stack
What it runs on.
- Warehouse
- Amazon Redshift
- Pipeline
- AWS Managed Airflow, EMR Serverless, Amazon S3
- Sources
- Sybase, Strapi, MySQL, Google Sheets-300+ tables
- Visualisation
- Apache Superset, 15+ dashboards
The road ahead
The sources are joined. Next is what that makes answerable.
With a centralised warehouse and a self-serve pipeline, adding a source or a dashboard is configuration rather than a project-and ML models can run across data that previously could not be combined.
- 01
ML models across joined sources
- 02
More dashboards per business line
- 03
Inventory optimisation on live data
Talk to us
Bring one process. Leave with an outcome architecture.
30 minutes on one function, mapped to the architecture behind it. No slides.