Aurora Insights
Multi-tenant SaaS analytics console
Our reference architecture for multi-tenant analytics: isolated tenant data, cached aggregate queries, and a dashboard builder that renders sub-second on cold cache.
Built by us and demoable end to end.
Overview
Live workspace · updated just now
MRR
$48.2k
+12.4%
Active tenants
312
+18
Churn
1.8%
-0.4%
Revenue trend
Cached aggregates
Query path
Row-level
Tenant isolation
Config-driven
Widget types
Built in
Metering
The problem
Most early SaaS teams bolt reporting onto their transactional database, then watch a single customer’s dashboard take the whole product down.
Built with
- Next.js
- Postgres + RLS
- Redis cache
- Recharts
- Stripe metering
- Vercel
The approach
- 1
Tenant isolation at the row level, with every query forced through a tenant-scoped policy rather than trusted application code.
- 2
Read path split from the write path: aggregates are materialised on a schedule and served from cache, so dashboards never contend with production writes.
- 3
A widget contract (query + shape + formatter) so new charts are configuration, not new code.
- 4
Usage metering and plan gating designed in from day one instead of retrofitted at the first pricing change.
Shipped features
- Role-aware dashboards with per-tenant theming
- Scheduled digests and threshold alerts
- CSV / API export with rate limiting
- Audit log on every data access
Build something like this
Tell us the problem, not the feature list.
We start with the workflow and the money, then design the software that fits it.