Work / Meridian Analytics
SaaS / Data·App Development

Real-time analytics dashboard for enterprise clients

Key result
50msavg. query response time
Timeline: 6 weeksClient: Meridian Analytics
Meridian Analytics
12
Enterprise clients onboarded
99.9%
Uptime since launch
₹1.2Cr
New ARR in 6 months
01

The Challenge

Meridian was losing deals because their analytics were delivered as static PDF reports with a 48-hour delay. Enterprise clients demanded real-time, interactive dashboards they could embed in their own products. The internal team had attempted a build twice before — both stalled at the scaling problem (multi-tenant data isolation + sub-second query times on millions of rows).

02

What We Built

We built a multi-tenant dashboard platform with real-time WebSocket updates, custom D3.js visualizations, role-based access control, and a white-label embed system. The architecture handles millions of events per day with sub-50ms query times thanks to PostgreSQL + Redis caching and pre-aggregated materialized views. White-label support means every enterprise client gets their own subdomain, branding, and OAuth — all from one codebase.

03

What we learned

1

Cache the aggregates

Live raw queries on millions of rows = no. Pre-aggregated views refreshed every 30s = yes. 50ms became achievable.

2

Embeds are auth problems

Most of the engineering was authorizing embeds. Two weeks on tokens, JWT scopes, and signed URLs.

3

Multi-tenant from day 1

Adding tenancy retroactively is a nightmare. We baked it into the schema from the first migration.

Stack

Technologies we used

ReactNode.jsPostgreSQLRedisAWSD3.jsTypeScript

We came in with a vague idea of what we wanted. They shipped a real, working analytics platform our enterprise clients now embed in their own products.

AS
Anya Sharma
CEO, Meridian Analytics

Have an idea? Let’s build it.

Tell us what you want to build. We respond within 24 hours.

Start your project
Real-time analytics dashboard for enterprise clients — Kiteways