Business Context
A high-volume enterprise relied on a complex network of 50+ interlinked Google Sheets to track revenue, expenses, and inventory across nine different sub-brands. As the database grew, "spreadsheet lock" and data inconsistency became critical risks to daily operations.
Problem
Lack of a "Single Source of Truth." Financial totals in the marketing sheet didn't match the inventory sheet. Month-end closing took 7 days of manual reconciliation. The risk of data corruption was highest during multi-user access periods.
Constraints
- Non-technical staff must still be able to input data easily.
- Existing historical data (3 years) must be migrated without loss.
- Implementation must not interrupt daily operations.
Architecture Design
I moved the core data storage from Sheets to a professional PostgreSQL database hosted on Supabase. I designed a relational schema that enforced data types and foreign key relationships, making "orphan" data impossible.
Implementation
I built a bridge using Next-auth for secure access and a custom Vue.js dashboard for high-stakes data entry. To satisfy the need for spreadsheet-like interfaces, I integrated React-Table with direct Supabase bindings. Data validation was moved from the client-side to database-level triggers and constraints.
Business Impact
Scale
Successfully migrated and synchronized 100+ data tables from legacy sources.
Speed
Financial reporting time reduced from 7 days to near real-time dashboards.
Reliability
100% data consistency achieved across 9 brands using the Unified CRM system.
Insight
Real-time Looker Studio BI dashboards replaced static, error-prone manual reports.
Technical Stack
- Supabase
- PostgreSQL
- Docker
- Python (ETL)
- SQL Triggers
- Looker Studio