KR
KARAN.
BACK TO ALL WORK
PROJECT 03Data Analytics & Full-Stack2025

OmniMetrics

Enterprise POS & Financial Analytics Intelligence Hub

ReactNode.jsExpress.jsGoogle BigQueryPostgreSQLSQLTailwind CSSRechartsRedis Cache
omni-metrics.bi/pos
5M+ ROWS
POS Transactions1.8s Query
QUERY OPTIMIZED

25x Faster Execution

12+ METRIC VIEWS

React + Recharts

STACK: React • Node.js • Google BigQuery • PostgreSQLBIGQUERY BI
5M+ Rows
Records Processed
25x Faster
Query Speedup
12+ Live
Data Modules

Architectural Challenges

High-volume retail and restaurant chains generate millions of raw transactional data points daily that take minutes to query on traditional relational databases, delaying critical pricing and supply chain decisions.

Engineering Strategy

Architected a cloud data analytics pipeline connecting a responsive React executive portal to Node.js/Express query engines optimized for Google BigQuery and PostgreSQL, reducing dashboard query latency from 45s down to 1.8s.

VERIFIED SYSTEM FEATURES

5 MODULES

Connected live Google BigQuery data warehouse pipelines to stream, aggregate, and visualize 5M+ historical POS records.

Authored optimized analytical SQL queries utilizing partitioned and clustered tables for cost-effective BigQuery computing.

Developed 12+ executive interactive visualization modules with Recharts tracking hourly sales velocity, wastage patterns, and predictive inventory orders.

Implemented Redis multi-tier caching for sub-second delivery of recurring business intelligence reports.

Built automated export pipelines delivering scheduled CSV/PDF executive summaries to stakeholders.

TECHNICAL ARCHITECTURE & SCHEMA DECISIONS
01 •Google BigQuery partition and cluster schemas minimizing scanned byte volume per query.
02 •Node.js aggregation server with Redis cache key invalidation on new batch ingestion.
03 •React dashboard architecture with parallel query fetching and granular error boundaries.